XXXIst International Symposium on Technological Innovations in Laboratory Hematology


Printable Program




May 9-11, 2019

Vancouver, Canada




4:30 - 5:30 PMVirtual Room 2
PLENARY - ORAL ABSTRACTS

Best Student Presentation Award Finalist
PEripheral Blood Eryptosis And Its Correlations With Erythroblastic Apoptosis In The Bone Marrow In Hematological Disorders
Chander Hans1, Prashant Sharma1, Rahul Saini1, Man Updesh Singh Sachdeva1, Parveen Lata Bose1, Minakshi Gupta1, Alka Rani Khadwal2, Reena Das1, Neelam Varma1
1Hematology Department, Postgraduate Institute of Medical Eduation and Research (PGIMER), Chandigarh, India, Chandigarh, India, 2Adult Clinical Hematology and BMT Unit, PGIMER, Chandigarh, India

Introduction: Eryptosis, the suicidal death of erythrocytes, has been extensively studied in physiological and pathological states. It’s relationship however, with apoptosis of erythroid precursors remains unexplored, despite their several shared triggers and pathogenetic mechanisms.
Methods: We evaluated peripheral blood eryptosis in 53 patients undergoing bone marrow aspiration for various hematological disorders. Flow cytometric assessment of phosphatidylserine exposure by annexin V binding and calcium influx by fluo-3AM was done along with measurement of side scatter [Figure 1 shows the gating strategy for annexin V binding on erythrocytes. Singlet gating/doublet discrimination on FSC-area versus height was followed by gating of erythrocytes on log forward-versus-side scatter. Interval gates were used to define erythrocyte populations on FITC-versus-cell number histograms.] Early and late apoptosis of CD71+ve erythroblasts and presence of dyserythropoiesis on marrow smears were determined within corresponding bone marrow aspirates.
Results: Median age was 32 years (range 1-75 years); 38 (72%) had Hb ≤11.0 gm%. Patients overall had significantly higher annexin V-binding (phosphatidylserine exposure) and fluo-3AM signal (calcium influx) and lower MCV-adjusted SSC vis-à-vis controls. Annexin V-binding and fluo-3AM fluorescence correlated strongly with each other (r=0.885, p< .001). Moderate correlations were observed for MCV-adjusted SSC with both annexin V and fluo-3AM (r=0.613 and 0.596 and p=.042 and .017 respectively). Phosphatidylserine exposure and Ca2+ influx were increased in 64% and 36% cases respectively. The 15 (28%) patients with increased annexin V as-well-as fluo-3AM-binding had significantly lower hemoglobins and reticulocyte counts and increased red cell distribution width and circulating nucleated RBC numbers vis-à-vis the others. Dyserythropoiesis correlated significantly with erythrocytic phosphatidylserine exposure (r=.618, p=.014) but not with fluo-3AM-binding (p=.262). Patients with dyserythropoiesis had significantly higher early apoptosis compared to those without dyserythropoiesis (p=.006). Late apoptosis correlated significantly with hemoglobin (p=.009) and erythrocytic phosphatidylserine exposure (p< .001). Figure 2 shows illustrative cases.
Conclusions: This is the first study to compare and demonstrate links between peripheral blood eryptosis and erythroblastic apoptosis. Establishment of eryptosis and apoptosis’ interrelationships in patients with diverse disorders connects the marrow environment to peripheral blood and reveals eryptosis in previously unrecognized conditions.

Best Student Presentation Award Finalist
Factor Viii Inhibitor TItres Are Consistently Lower For Chromogenic Compared To One Stage Factor Viii Assays - Carol Briggs Top Scoring Abstract
Karen A. Moffat1,2, Stephen Carlino1, Catherine P.M. Hayward1,2,3
1Hamilton Regional Laboratory Medicine Program, Hamilton, ON, Canada, 2Department of Medicine, McMaster University, Hamilton, ON, Canada, 3Department of Pathology and Molecular Medicine, McMaster University, Hamilton, ON, Canada

Introduction: With the launch of new therapies, such as Emicizumab, laboratories may be required to offer testing for factor VIII (FVIII) inhibitors by APTT or chromogenic assays. We undertook a comparison of these methods to evaluate potential discrepancies in estimates of FVIII inhibitor levels.
Methods: FVIII inhibitor titres were determined by a modified Nijmegen method, using one stage, APTT-based FVIII (Siemens Actin FS, Marburg, Germany) and Chromogenix COAMATIC Factor VIII (that uses bovine FIXa; Instrument Laboratories, Bedford, MA) assays to quantify residual FVIII. Sodium citrate plasma samples were stored at –70oC and thawed (5 minutes, 37oC waterbath) immediately prior to inhibitor testing.  Samples evaluated included: a FVIII inhibitor quality control (QC) (Affinity Biologicals, Ancaster, ON), run neat and diluted (respective established QC ranges for APTT-based FVIII inhibitor titres: 5.6-7.6 and 1.0-2.0 BU); hemophilia A (inherited: n=6; acquired: n=2) patient plasma; and two ECAT external quality assessment (EQA) samples. Results were analyzed by paired T-tests (p< 0.05 considered significant) and Bland Altman plots.
Results: There was complete agreement between the assays for samples with an undetectable FVIII inhibitor (< 0.5 BU) (Table 1). All other samples had significantly lower FVIII inhibitor titres estimated by the chromogenic compared to the APTT-based FVIII assay (p < 0.001; Table 1). Bland Altman plots demonstrated negative bias for FVIII inhibitor titres estimated by the chromogenic FVIII assay, and while differences in the titres were modest for some samples (1.2-1.9 fold differences), they were striking for others (>15 fold difference).
Conclusions: Differences in FVIII inhibitors titres, estimated by APTT-based versus chromogenic FVIII assays, could complicate the monitoring of FVIII inhibitor levels in patients with hemophilia A. The use of bovine FIXa in the chromogenic assay may underlie the underestimation of human antibodies against FVIII in hemophilia plasma. Laboratories and clinicians should be aware of these clinically important differences, which have implications for ongoing monitoring of FVIII inhibitor titres. To minimize risks for individual patients, we recommend performing a baseline assessment of FVIII inhibitor titres, using chromogenic FVIII assays with bovine FIXa, for hemophilia A patients that will receive Emicizumab or that need a new baseline inhibitor titre established while on Emicizumab.

The Effect Of Image Data Augmentation For Artificial Intelligence Model In Mature Leukocyte Classification
Hiroyuki Nozaka1, Honoka Harako2, Mae Miyazaki2, Suzuka Kaga2, Niina Sakaiya2, Kyouka Kudo2, Shou Kimura2, Manabu Nakano1, Miyuki Fujioka1, Kazufumi Yamagata1
1Hirosaki University Graduate school of Health Sciences, Hirosaki, Japan, 2Hirosaki University School of Health Sciences, Hirosaki, Japan

Introduction: Artificial Intelligence (AI) is rapidly developing in recent years as a technology that enables judgments that reflect thought patterns. The deep learning automatically extracts the common features on the training images, and it realizes accurate and efficient judgment. Although it is expected to automate blood morphology analysis by the deep learning, a lot of training images are required for optimal AI model creation. The data augmentation is known as one of the techniques to create high quality and sufficient amount of training data set from a small data set. However, invalid data augmentation is also reported in recent AI study, there are few reports on optimal AI model creation techniques in hematology tests. Therefore, we performed blood morphology analysis with the deep learning, and evaluated usefulness of data augmentation in leukocyte classification.
Methods: The original image data set for training was consisted of 1335 typical mature leukocyte images. Leukocyte images were classified into six categories. Data augmentation process (Rotation/Distortion/Scaling) was applied to the original image. ResNet-18/34/50 model were used for deep learning, transfer learning and fine tuning for optimization was performed. Nnabla (SONY.INC) was used as the deep learning library. The subjects for evaluation were peripheral blood smear of 51 healthy persons.
Results: The accuracy of AI model for leukocyte 6 classification trained with rotation images showed 0.855 with ResNet-18, 0.807 with ResNet-34 and 0.845 with ResNet-50. The depth of the learning layer did not show any improvement in leukocyte classification accuracy. The accuracy of AI model trained with scaling images showed 0.835 with scale ratio 5%, 0.855 with scale ratio 20%. Although segments and bands showed improved accuracy, lymph and mono did not show improved accuracy (Table1, 2).
Conclusions: It was cleared that image data augmentation with rotation contributed to the improvement of leukocyte classification accuracy. In contrast, the effect of data augmentation with distortion or scaling is limited to specific cell categories in improving accuracy. Although data augmentation is effective in generating blood morphology AI models with the deep learning, it is necessary to select the data augmentation process to be applied for mononuclear or segmented cells.

Best Student Presentation Award Finalist
Adaptive Natural Killer Cells Confer Innate Protection In Covid-19 Infection
Vincent, Yi Sheng Oei1, Guat Bee Tan2, Stephrene Seok Wei Chan1, Dheepa Christopher1, Gek Hsiang Lim3, Kiat Hoe Ong1, Bingwen Eugene Fan1
1Department of Haematology, Tan Tock Seng Hospital (TTSH), Singapore, Singapore, 2Department of Laboratory Medicine, TTSH, Singapore, Singapore, 3CRIO, TTSH, Singapore , Singapore

Introduction: Natural killer (NK) cells are innate lymphocytes that exhibit swift, cytotoxic anti-viral response. NKG2C+/NKG2A-/CD57+ Adaptive NK are more prevalent in critically-ill COVID-19 patients, but no studies were performed on their role in asymptomatic COVID-19 patients.
Methods: We studied the NK cell subsets of 55 patients admitted to the National Centre for Infectious Disease(Singapore) between August to October 2020 by performing peripheral blood 8-colour flow cytometry(CD45-APC-H7/CD3-AF700/CD19-BV605/CD56-PE-Cy7/NKG2A-BV421/NKG2C-BV480/CD4-FITC/CD57-APC). Results were correlated with clinical data including patient demographics, PCR Cycle Threshold (PCR CT) values from nasopharyngeal swabs on admission, symptoms, laboratory and radiologic investigations. 
Results: 31 patients were South Asians. 27 patients were asymptomatic with higher PCR cycle-threshold value, higher WBC count, lower CRP, and absent pneumonia on Chest X-ray (CXR). Flow cytometry analysis revealed significantly (p< 0.01) higher proportion of asymptomatic individuals with adaptive NK expansion (63%) (Table 1, Figure 1). Besides increased adaptive NK cells, asymptomatic individuals also have higher percentage of NKG2C+ cells but lower percentage of CD57+/NKG2C- cells. Furthermore, adaptive NK expansions are commoner in south (16/31) or southeast (7/14) Asians comprising almost 50% of each subset compared to 20% reported in Europe.
  Overall Asymptomatic Symptomatic P-value
Males Females n(%) 30(54.6) 25(45.5) 10(37.0) 17(63.0) 20(71.4) 8(28.6) 0.01
PCR CT 26.3(15.5, 41.5) 32.4(16.3, 41.5) 24.7(15.5, 25) 0.004
WBC(x 109) 7.4(2.8-16.9) 9.3(4.9-16.9) 6.3(2.8-14.2) 0.0001
CRP(mg/L): 3.8(0.3-214) 2.2(0.3-36.8) 6.6(0.3-214) 0.01
CXR; n(%) Clear;Abnormal;  Pneumonia   44(80.0);4(7.3); 7(12.7)   25(92.6);2(7.4);            0(0.0)   19(67.9);2(7.1);          7(25.0) 0.01
Adaptive NK expansion:n(%) No, Yes 31(56.4),    24(43.6)  10(37.0),             17(63.0) 21(75.0),           7(25.0) 0.005
NKG2C+ NK:%total NK 0.16(0.02-0.64) 0.33(0.02-0.62) 0.07(0.02-0.64) 0.0007
Adaptive NK:% total NK 0.06(0.006-0.59) 0.21(0.007-0.50) 0.04(0.006-0.59) 0.0009
Table 1: Significant clinical and laboratory parameters of NK clinical study
Conclusions: COVID-19 patients with adaptive NK cells (NKG2C+ subtype) could likely promptly eliminate SARS-CoV-2 infected cells, leading to asymptomatic infections. This correlated with lower CRP, greater CT-values, lesser CXR pneumonia and higher-normal leukocyte counts, with a significant proportion of South Asian COVID-19 patients found to carry the adaptive NK cell expansion.

5:45 - 6:45 PMVirtual Room 1
ORAL ABSTRACTS 1: Cellular Analysis/Morphology

Best Student Presentation Award Finalist
Evaluation Of Kinetics And Single Cell Sequencing Based Phenotype Of Leukemic Stem Cells (Lscs) In Acute Myeloid Leukemia
Ritu Gupta1, Nupur Das1, Deepshi Thakral1, Nitu Jha1, Vivek Singh1, Saroj Singh1, Sandeep Rai1, Vijay Kumar Prajapati1, Sameer Bakhshi2, Ranjit Kumar Sahoo2, Ajay Gogia2, Atul Sharma2, Lalit Kumar2
1Laboratory Oncology, All India Institute of Medical Sciences (AIIMS), New Delhi, India, 2Medical Oncology, All India Institute of Medical Sciences, New Delhi, India

Introduction: Leukemic stem cells (LSCs) possess leukomogenic and drug resistance properties but their clinical significance is yet to defined. We assessed the frequency of LSCs and correlated it with treatment response in AML patients. Abseq based single cell sequencing (scSeq) was used to assess the efficiency of immunophenotypic markers  in discriminating LSC from HSC. 
Methods: Bone marrow aspirates from 190 AML patients treated with 7+3 induction chemotherapy were processed for immunophenotyping where LSCs were characterized as CD34++CD38-CD45 dim population, expressing CD123 or CD33 or CD11b. Another tube with CD44, Combi6 and CD45RA was also used to characterize LSC. A sample was labelled MRD+ with presence of  ≥0.1% blasts with aberrant phenotype. For scSeq, CD34+CD38- populations sorted from leukemic (LSC; n=3) and control bone marrows (HSC; n=2) were labelled with Ab-Oligos (n=19) followed by single cell capture, cDNA synthesis, library preparation and paired end sequencing on NovaSeq6000. The flow cytometric and scSeq data were analyzed using kaluza V2.0 and Seven Bridges software, respectively.
Results: 65.3% achieved morphological remission (CR) and 58% attained MRD negativity. The non-CR and MRD+ groups were associated with higher numbers of LSCs (Table 1). The relative expression profile of Abseq markers revealed higher numbers of CD34 and CD45RA molecules on LSC compared to HSC (Figure 1).  Table 1:  Frequency of LSC at diagnostic and post induction time-points and response to induction therapy
Response at post induction LSC numbers at diagnosis median (range) p value LSC numbers post induction median (range) p value
CR 0.1480 (0 -7.974) 0.001 0.025 (0-1.005) 0.68
Non-CR 0.510 (0–8.07) 0.074 (0–6.702)
MRD+ 0.413 (0.022–8.072) 0.007 0.075 (0-6.702) 0.002
MRD- 0.140 (0–7.074) 0.0259 (0- 0.828)

Conclusions: The treatment response is inferior in patients with higher LSCs possibly because the induction regimen is targeted towards tumour bulk and not LSCs. The single cell sequencing data supports use of CD34 and CD45RA to discern LSC from HSC which may be useful in ascertaining the residual normal HSC compartment in AML patients.

Evaluation Of The Scopio Labs X100 Full Field Pbs System &Ndash; A Novel Digital Imaging System That Utilizes A Full Field View Approach For The Comprehensive Analysis Of Peripheral Blood Smears
Ben-Zion Katz1,2, Michael D. Feldman3, Minychel Tessema4, Dan Benisty1, Grace Stewart Toles3, Alicia Andre4, Bronka Shtreker1, Maria Fatima Paz3, Joshua Edwards3, Darrin Jengehino3, Adam Bagg3, Irit Avivi1,2, Olga Olga Pozdnyakova4
1Tel Aviv Medical Center, Tel Aviv, Israel, 2Sackler Faculty of MedicineTel Aviv University, Tel Aviv, Israel, 3Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA, United States, 4Department of Pathology, Brigham and Women's Hospital, Boston, MA, United States

Introduction: Current digital cell imaging systems preform analysis in limited regions of the peripheral blood smear (PBS), and may require the support of manual microscopy without achieving full digital workflow. This study evaluated the performance of the FDA cleared Scopio Labs X100 Full Field PBS system, which provides high-resolution full field viewing of PBS, combined with AI-based pre-classification of leukocytes and quantitative platelet estimation.
Methods: We analyzed 335 normal, and 310 abnormal PBS from patients with various infectious or neoplastic conditions. WBC differentials, RBC morphology and platelet estimation were performed at three sites, using Scopio’s Full Field PBS as the test arm, and manual microscopy as the reference arm. Accuracy, sensitivity and specificity were compared based on the CLSI H20-A2.
Results: Normal ranges were comparable between the two methods and between the three sites. The accuracy of the morphological abnormalities, distributional abnormalities, and total accuracy were 96.82%, 95.75% and 96.29%, respectively. The sensitivities of the same groups were 85.46%, 88.83% and 87.86%, and the specificities were 97.79%, 97.43% and 97.62%, respectively. The overall agreement between the test and reference method for the RBC morphology was 99.77%, and the comparison for platelet estimation resulted in an accuracy of 94.89%, sensitivity of 90.00% and specificity of 96.28%, and no significant bias. High levels of repeatability and reproducibility were recorded for both WBC and platelets. The full-field viewing capability was demonstrated on a variety of clinical conditions associated with distributional and morphological abnormalities (for example, acute leukemia https://demo.scopiolabs.com/#/view_scan/8bad8d9d-4260-45ca-a691-82e2b9d904a0) 
Conclusions: Scopio’s Full Field PBS showed an excellent performance with a high degree of correlation with manual microscopy. The novel full field view of specimens enables a full digital PBS review and an anticipated disengagement from the manual microscope. 

Best Student Presentation Award Finalist
Survival Analysis Of Sars-Cov-2 Infected Patients Using Blood Cell Count And Morphology Parameters
Javier Nieto-Moragas1,2, Anna Marull1,2, Meritxell Deulofeu1,2, Orlando Jimenez1,2, Francisco Reina2, Raúl Benitez3, Maite Serrando1,2
1Laboratori Clínic Territorial de Girona. Parc Hospitalari Martí i Julia. , Girona, Spain, 2Departamento de Ciencias Médicas. Grupo de Investigación en Anatomía Clínica, Embriologíay Neurociencia. Universitat de Girona., Girona, Spain, 3Centre de Recerca en Enginyeria Biomèdica, Escola d’Enginyeria de Barcelona Est, Universidad Politécnica de Catalunya., Girona, Spain

Introduction: The study of the cells present in peripheral blood may help to differentiate patients with poor prognosis. Hematology analyzers report morphometric data called cell population data (CPD) that correlate with the findings during microscopic examination of the peripheral blood smear. The aim of the study is to determine the usefulness of cell population data and other laboratory parameters for the prognosis of SARS-CoV-2 infected patients.
Methods: 119 patients attending the emergency department in the period March-April 2020 were studied. The study inclusion criterion was a positive result for SARS-CoV-2 by real-time polymerase chain reaction (RT-PCR) and to determine laboratory parameters. Mean comparison of the analytical results between the survivor and exitus groups was performed using Mann-Whitney test. A Kaplan-Meier analysis was used to determine the significance when prognosis is studied.
Results: 119 patients with SARS-CoV-2 infection during hospital admission were studied, of whom 24 were exitus (mortality rate: 20%). Significant differences were only observed between the two groups for CRP and leukocyte subpopulation counts (Table 1).  
Parameter Survival group Exitus group p-value
C reactive protein (mg/dl) (CRP) 6.48 [9.5] 17.05 [12.48] < 0.001
Neutrophils count (103/µL) 4.64 [2.91] 8.41 [6.03] < 0.001
Lymphocytes count (103/µL) 0.96 [0.57] 0.65 [0.51] < 0.01
Table 1. Meand and interquartile range between survival and exitus group.  NE-WY (fluorescent intensity width, correlated with genetic material) showed significant differences (p< 0.001). The Kaplan Meier analysis showed a high discriminative power of the N/L ratio and the morphometric parameter NE-WY (table 2) to differentiate between survival and exitus groups.    
Parameter Survival group Exitus group Cut-off Log-rank test p-value
C reactive protein (mg/dl) 6.48 [9.5] 17.05 [12.48] 15 0.015
N/L ratio (arbitrary units) 5.0 [4.12] 10.05 [9.18] 7 < 0.001
NE-WY (arbitrary units) 622.0 [56.0] 669.5 [89.75] 660 < 0.001
Table 2. Kaplan-Meier analysis between survival and exitus group.
Conclusions: Leukocyte subpopulations count together with morphological characteristics determined with hematological analyzers can help to differentiate unfavorable prognoses in order to rapidly implement therapeutic actions during hospital admission.

Best Student Presentation Award Finalist
A Machine Learning Tool Using Digital Microscopy (Morphogo) For The Identification Ofabnormal Lymphocytes In Bone Marrow
Gusheng Tang1, Xinyan Fu2, Guoping Zhang3, Zhen Wang4, Mingyi Chen5
1Department of Hematology, Changhai Hospital, Shanghai, China, 2Division of Medical Technology Development, Hangzhou Zhiwei Information & Technology Ltd., Hangzhou, China, 3Department of Hematology, Xiangya Hospital of Central South University, Changsha, China, 4Clinical laboratory, The First Affiliated Hospital of Dalian Medical University, Dalian, China, 5Department of Hematopathology, UT Southwestern Medical Center, Dallas, TX, United States

Introduction: Morphological analysis of bone marrow is an essential step in the diagnosis of hematological diseases. The analysis of bone marrow smears by manual count is labor-intensive and subject to inter-observer variability. The morphological differential diagnosis of abnormal lymphocytes from normal lymphocytes is still challenging. The digital pathology methods integrated with advances in machine learning enablethe development ofnew diagnostic features/algorithms on digital cell images in order to optimize classification, thus providing a robust, and faster screening diagnostic tool.
Methods: We have developed a system Morphogo based on machine learning algorithms to discriminate abnormal lymphocytefrom normal lymphocytes using digital imaging analysis. We retrospectively reviewed 347 cases of bone marrow with high resolution images. Among them, 53 cases had clinical history and were diagnosedto have marrow involvement with lymphoma either by morphology or by flow cytometry.  Wesplit the 53 cases into two groups for training and testing datasets with 43 and 10 cases respectively. The selected 15,353 images were reviewed by pathologists, based on morphologic visual appearance, from the 43 patients, in training group,whose diagnosis was confirmed by complementary tests. To expand the range and the precision of recognizing the lymphoid cells in the marrow by the system, we developed an algorithm that incorporated cell size, nuclear/cytoplasmic color and texture in addition to geometrical cytologic features of the variable lymphocytes images,which were applied as the training dataset.
Results: The selected images from the 10 patients were analyzed by the trained AI based recognition system and compared with the final diagnosis rendered by pathologists.  The positive predictive value for the identification of the categories of reactive/normal lymphocytes and abnormal lymphoid cells, was 99.04%.  It seems likely that further training and improvement of the algorithms will facilitate further subclassification of specific lineage subset pathology, e.g., diffuse large B-Cell lymphoma from chronic lymphocytic leukemia/small lymphocytic lymphoma, follicular lymphoma, mantle cell lymphoma in the cases of abnormal malignant lymphocyte classes in future. 
Conclusions: This study demonstrated the feasibility of digital pathology and emerging machine learning approaches to automatically diagnose lymphoma cells in the bone marrow based on cytological-histological analyses.

5:45 - 6:45 PMVirtual Room 2
ORAL ABSTRACTS 2: Coagulation

Best Student Presentation Award Finalist
Significance Of Hematologic And Coagulation Parameters In Patients With Covid-19 
Narendra Bhattarai, Meryem Terzioglu, Frido Bruehl, Eleanor Cook, Megan Nakashima, Kandice Kottke-Marchant, Heesun Rogers
Cleveland Clinic, Cleveland, OH, United States

Introduction: COVID-19 can affect hematologic and hemostasis parameters resulting in hypercoagulability and poor outcomes. Antiphospholipid antibodies (APAs) are associated with thrombosis and can arise transiently during critical illness. The literature is mixed on whether or not APAs contribute to COVID19-associated thrombosis. This study aims to describe the clinical significance of hematologic and coagulation parameters including APAs in COVID-19 patients.
Methods: Clinical and laboratory data of patients admitted for COVID-19 infection with lupus anticoagulants (LAC) diagnostic panels ordered in 03/2020-09/2020 were retrospectively reviewed.
Results: 112 COVID-19 patients (mean 65 years; 51 male, 61 female) were included. Interim between admission and COVID-19 tests and between COVID-19 and LAC tests were median 0 and 2 days, respectively. Table/Figure presents detailed results. COVID-19 patients showed leukocytosis (28%), anemia (47%), thrombocytopenia (16%), and lymphopenia (59%) with increased neutrophil/lymphocyte ratio (8.6). APTT, D-dimer, fibrinogen, factor VIII, C-reactive protein(CRP) and ferritin levels were elevated. LAC was cancelled in 8 patients who received high heparin or anti-Xa drugs.  LAC was positive in 34.6%, indeterminate in 26.9% and negative in 38.5%. LAC-positive patients had significantly elevated APTT and CRP, and decreased lymphocytes compared to LAC-negative patients (p< 0.05). In LAC-positive patients, phospholipid-dependent tests showed positivity in STACLOT 89.5%, DRVVT 34.2% and PNP 21.1%. CRP was significantly higher in STACLOT-positive patients than STACLOT-negative patients (13.4 vs 6.5 mg/dL, p< 0.001), however, there was no difference in CRP based on DRVVT (9.0 vs 10.0 mg/dL, p>0.05). In LAC-positive patients, DRVVT mixing was positive in 60.5% and APTT mixing 78.9%. Anticardiolipin antibody(ACA) and B2-glocoprotein-1(B2GP1) were positive in 7.3% (5IgM, 2IgG, 1IgA) and 6.4% (5IgM, 2IgG). 9 cases were double/triple positive. Repeat LAC tests (N=7; median interim 40 days) were indeterminate or negative. 12.5% developed thromboses (8 deep vein thrombosis, 4 pulmonary emboli), 32.1% were admitted to ICU, and 18.8% expired. Sepsis/respiratory failure was significantly higher among LAC-positive patients (45.0%) than LAC-indeterminate/negative patients (14%)(p< 0.05). Thrombosis was more common among LAC-positive patients (20.0%) than LAC- indeterminate/negative patients (8.0%)(p=0.06). Thrombosis was similar in ACA/B2GP1-positive or negative patients (10% vs 12%, p>0.05).
Conclusions: COVID-19 infection in admitted patients was associated with hypercoagulability and unfavorable outcome due to acute infection-related multifactorial consequences. LAC positivity was associated with lymphopenia, elevated CRP and high sepsis/respiratory failure in this cohort.

Effect Of Doac-Filter On Thrombophilia Coagulation Parameters
Eleni Linskens, Katrien Devreese
Coagulation Laboratory, Department of Laboratory Medicine, Ghent University Hospital, Ghent, Belgium

Introduction: Direct oral anticoagulant (DOAC) interference on measurement of thrombophilia parameters can be prevented by various strategies, among which the recently available DOAC-Filter device (Stago). We aim to evaluate the effect of DOAC-Filter on routine thrombophilia coagulation parameters. 
Methods: FVIII activity (STA-C.K. PREST, STA FVIII-deficient plasma [Stago]), protein C activity (PC,  Coamatic Protein C assay [Chromogenix, Werfen-IL]), Free Protein S antigen (PS, STA-Liatest Free Protein S reagent [Diagnostica Stago]) and antithrombin (AT,  BIOPHEN AT anti-(h)-Xa LRT [Hyphen Biomed]) were measured before and after DOAC-Filter treatment (Stago; used according to manufacturer’s recommendations) on STA-R Evolution® analyzer (Stago). Three normal pool plasma (NPP) samples, twenty normal patient samples and 30 deficient patient samples (AT, PC, PS n=10 each) were included. Mean percentage differences (NPP and normal samples) and 95%CI (normal samples) were calculated. Differences before and after DOAC-Filter treatment were regarded clinically significant when exceeding the acceptable bias according to CLSI-C56-A guidelines: “1.96x(CVa2+CVw2)1/2” (NPP CVa=CV NPP measurements, patient CVa=in-house established CV, CVw=within-subject biological variation [Westgard database]). Wilcoxon-signed-rank p-values were calculated (Medcalc®). Deficient patient sample results before and after DOAC-Filter were compared to in-house reference ranges. 
Results: NPP results showed significant differences, except for FVIII. In normal patient samples a statistically significant FVIII decrease (-9.9%) after DOAC-Filter treatment was observed, however not clinically significant. AT, PC and PS showed a clinical significant increase (Table 2). AT, PC and PS deficient patient samples ranged from 24%-57%, 28%-54% and 13%-61%, respectively, showing mean percentage differences of 14.5%, 4.1% and -1.8% after DOAC-Filter treatment respectively (not clinically relevant). For PC and PS, no difference in interpretation was seen according to the reference ranges for all 10 samples. One AT deficient sample showed an increase (from 57% to 76%) leading to an AT level within normal ranges after DOAC-Filter treatment.
Conclusions: A significant increase is seen for AT, PC and PS after DOAC-Filter treatment in normal samples, but with no obvious change in clinical decision making except on AT for one patient sample.The procoagulant effect of DOAC removing agents have been shown before. We illustrated here the increase of thrombophilia parameters, confirming not to use DOAC removing agents/devices for samples not containing DOAC. DOAC-Filter deserves further evaluation on samples from patients treated with DOAC. Also, further evaluation of a level-dependent effect of DOAC-Filter is needed.

Platelet And Endothelial Dysfunction Complex For Gestational Arterial Hypertension During Pregnancy After In Vitro Fertilization
Victoria Soboleva1, Alla Shabalina2, Eugene Roitman2, Leonid Alexandrov1
1Sechenov First Moscow State Medical University, Moscow, Russia, 2Research Center of Neurology, Moscow, Russia

Introduction: Hemocoagulation and endothelial disorders play a role in gestational arterial hypertension (GAH) pathogenesis, which can lead to preeclampsia, and gestational diabetes, and cardiovascular and cerebrovascular complications. The aim was to reveal how the GAH is dependent on platelet and endothelial conditions in pregnancy occurred after in vitro fertilization (IVF) with oocyte donation.
Methods: The study included 40 patients (age 46.1±4.8), of whom 22 patients developed the GAH in the third trimester (group I) and 18 pregnant women were without the GAH (group II). Laboratory testing was performed in every trimesters with ADP-, and collagen-, and epinephrine-platelet aggregation, plasminogen, alpha-2 antiplasmin, tissue plasminogen activator (tPA), tissue plasminogen activator inhibitor, Willebrand factor (vWF), ADAMTS-13 and soluble thrombomodulin (sТМ) by conventional methods. Statistical processing was carried out with descriptive statistics, nonparametric methods using Mann-Whitney and Fisher criteria, and multivariate analysis of variance.
Results: In both groups we found remarkable platelet activation and endothelial dysfunction throughout the gestational period starting from the I trimester. However, in II and III trimesters the platelet aggregation was lower compared to the I trimester that was obviously as a result of antiplatelet administration. Group I showed greater values of platelet aggregation, vWF and sTM together with smaller values of tPA, and ADAMTS-13 compared to Group II throughout the full gestational period. The GAH development in III trimester correlated with some parameters measured in I trimester: ADP-induced platelet aggregation (r=0.321, p< 0,01), collagen-induced platelet aggregation (r=0.334, p< 0.01), ADAMTS-13 (r=-0.326, p< 0.01), tPA (r=-0.287, p< 0.05), sТМ (r=0.283, p< 0.05). Despite of weak relations separately, modeled integral parameter joined them in one has presented r=0.68 (p< 0.01) for the GAH development in III trimester.
Conclusions: In III trimester the GAH development in patients undergoing in vitro fertilization (IVF) with oocyte donation is associated with not only separate platelet and endothelial disorders but mainly with a complex of platelet and endothelial interactions in I trimester.

Performance Of A New Fully-Automated Chemiluminescent Assay For Adamts13 Activity For The Rapid Diagnosis Of Thrombotic Thrombocytopenic Purpura
Pierre Toulon1, Ines Harzallah2
1CHU Hopital Pasteur, Nice , France, 2CHR Muller, Mulhouse, France

Introduction: Thrombotic thrombocytopenic purpura (TTP) is a rare thrombotic microangiopathy (TMA) that can be caused by congenital or acquired severe deficiency of a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13 (ADAMTS13), an enzyme that cleaves the von Willebrand factor (VWF). ADAMTS13 deficiency leads to the accumulation of ultra-large VWF multimers, which causes platelet aggregation and the formation of microthrombi. So beside clinical symptoms, diagnosis of TTP laboratory usually include thrombocytopenia, hemolytic anemia, the presence of schistocytes and a severe reduced activity of ADAMTS13 (< 10%). As TTP is a life-threatening condition characterized by microvascular thrombosis, it is of great importance to shorten the time to the treatment. In that condition, a rapid ADAMTS13 testing is crucial for an early diagnosis and optimal management of acute TTP. The aim of the study was to evaluate the performance of a new fully-automated assay for ADAMTS13 activity and to compare test results to those obtained using the FRETS-VWF73 assay.
Methods: The HemosIL AcuStar ADAMTS13 assay (Instrumentation Laboratory, IL, Bedford, MA, USA) is a quantitative activity assay for ADAMTS13 that is fully-automated on the ACL AcuStar chemiluminescent analyzer (IL). It is ready-to-use, cartridge-based and available on-demand with a turn-around time of 33 min, 24/7. We evaluated 46 samples (number to be extended) obtained from patients with acute TMA. Diagnosis of TTP was confirmed in 5 cases. Agreement between methods was assessed using the cut-off value of 10% that is the commonly used ADAMTS13 activity level allowing the diagnosis of TTP in case of deficiency.
Results: ADAMTS13 activity was not significantly different when evaluated using the HemosIL AcuStar ADAMTS13 and the FRETS-VWF73 assays [72.6% (< 0.2 126) vs. 79% (< 5  >150) ; p< 0.05] . Moreover the concordance of test results as whether the ADAMTS13 activity was below (n=5) or above 10% was found to be excellent (kappa=1.00) in the present series, allowing an accurate TTP diagnosis in the patients.
Conclusions: Together with its short turnaround time (33 min) and its full automation, our results suggest that the HemosIL AcuStar ADAMTS13 activity assay could be an assay of choice to rapidly measure ADAMTS13 activity in plasma and thus to establish the diagnosis of acute TTP in emergency settings.

5:45 - 6:45 PMVirtual Room 3
ORAL ABSTRACTS 3: Flow Cytometry

Best Student Presentation Award Finalist
Role Of Neutrophil Cd64, Monocyte Hladr And Sepsis Index As Biomarkers For Covid 19 Infection, Bacterial Sepsis And Other Systemic Inflammatory States.
Namrata Awasthi, Sridhar Mishra, Abhishek Bajpayee, Vandana Tiwari, Jyotsna Agarwal, Nuzhat Husain
Dr RMLIMS, LUCKNOW, India

Introduction: Upregulation of CD64 on neutrophils (nCD64) and downregulation of HLADR on monocytes (mHLADR) is known to occur in bacterial sepsis (BS) leading to high value of Sepsis index (SI=nCD64/mHLADRX100) which shows a trend towards greater mortality. Severe Covid 19 infection (SCOV19) with acute respiratory distress syndrome should be considered viral sepsis as per international definition of sepsis pertaining to organ dysfunction. Current study was undertaken to understand and assess role of above three parameters in cases of mild (MCOV19), SCOV19, systemic inflammatory response syndrome without infection (SIRS) and BS.

Methods: Study group comprised of healthy controls(HC) 15, SIRS 15, BS 22, MCOV19 17 and SCOV19 14 (total 83). nCD64 and mHLADR were quantitated as ABC (antibody bound per cell) by BD QuantibriteTM reagents as per manufacturer’s instruction on venous blood samples within 4 hours of collection on BD FACS CANTO flow cytometer and FACSDIVA software. Relevant laboratory investigations and clinical scoring to categorize patients into various groups as per standard guidelines were followed. None of the patients were on immunosuppressants till the time of sampling. Blood was collected within 12-48 hours of hospital admission.
Results: nCD64 was upregulated and highest in BS> SCOV19>SIRS>MCOV19>HC (difference in mean values highly significant, p< 0.001 except HC vs MCOV19 and SIRS vs SCOV19). mHLADR was downregulated maximally in both BS and SCOV19 followed by MCOV19>SIRS>HC (difference in means highly significant, p< 0.001 except SIRS vs MCOV19 and BS vs SCOV19). SI was highest in BS> SCOV19>SIRS>MCOV19>HC and difference in means were significant across all groups. Among SCOV19, 5/14 were non- survivors and all (100%) had SI value>72. nCD64< 1277 and SI< 48.85 suggest a diagnosis of SCOV19 over BS (nCD64= AUC 0.8149, p 0.001, sensitivity 71.43, specificity 77.27) (SI=AUC 0.7208, p 0.027, sensitivity 57.14, specificity 72.73). nCD64 and SI can discriminate MCOV19 and SCOV19 with 100% sensitivity and 70-94% specificity, p< 0.0001.
Conclusions: Significant downregulation of mHLADR in BS and SCOV19 suggest profound immunosuppression in both;discrimination is possible based on nCD64 and SI in an ICU setup. nCD64 and SI can serve as marker for triaging MCOV19 and SCOV19. Very high SI is associated with mortality in SCOV19.

Best Student Presentation Award Finalist
Association Of Aberrant Immunophenotype Of Plasma Cell Myeloma And Monoclonal Gammopathy Of Undetermined Significance With Laboratory Parameters
Sindhura Lakshmi Koulmane Laxminaraya, Indroneel Sen
Kasturba Medical College,Manipal, MAHE, Udupi, India

Introduction: Multiparametric flow cytometry (MFC) is useful in determining clonality of abnormal plasma cells as well as identifying aberrant phenotype in Plasma cell myeloma (PCM) and Monoclonal gammopathy of undetermined significance (MGUS). It plays a major role in therapy as well as evaluation of minimal residual disease along with other laboratory parameters and clinical features. We aimed to evaluate aberrantly expressed MFC markers and their association with laboratory parameters in cases of PCM and MGUS in comparison with normal polyclonal Plasma Cells (PC).
Methods: This is was a retrospective observational study of consecutives cases of suspected myeloma between July 2017 and June 2020. Demographic, clinical, and biochemical data was recorded. Flow cytometry files were reviewed for aberrant immunophenotype of plasma cells. The values were compared between groups of PCM/MGUS and those with reactive PC. Pattern of expression of CD19, CD56, CD117 in PCM, MGUS and PC were analysed for their correlation with haemoglobin, serum calcium, serum Immunoglobulin (Ig) and free light chain levels and clinical features. 
Results: Total 178 cases were included in the study (129 PCM, 16 MGUS and 33 reactive PC). M:F was 2.08:1, median age 63 years. Biochemical and CBC parameters are summarized in table 1. Pattern of expression of CD117, CD56, CD19 and cytoplasmic Kappa showed a significant difference between the 3 groups, and strong correlation with laboratory parameters (p < 0.001)(Table 2). Although no single variable predicted reactive PC, MGUS and PCM, a combination of Hb, RBC count, k/l ratio, S. creatinine, Blood urea and S. Calcium with CD19 negativity, CD56 and CD117 positivity provided significant descrimination between reactive PC versus MGUS and PCM.
Conclusions: Our study shows that MFC using CD117, CD19 and CD56 can reliably differentiate between PCM, MGUS and reactive PC. They also correlated well with biochemical parameters. MFC can accurately diagnose and monitor Plasma Cell disorders

Best Student Presentation Award Finalist
Phenotypic Alterations Of Hematogones In B-Lymphoblastic Leukemia Patients Treated With Immunotherapy
Weijie Li, Ruth E. Morgan, Roxanne Nieder, Sa T. Truong
Children's Mercy hospital, UMKC medical school, Kansas City, MO, United States

Introduction: Immunotherapy targeting leukemia antigens such as Inotuzumab Ozogamicin (IO)(anti-CD22 monoclonal–ozogamicin conjugate), Blinatumomab(CD19/CD3 bi-specific T-cell engager), Tisagenlecleucel (CART 19), Rituximab (CD20 monoclonal antibody) has been used in treating refractory or relapsed B-lymphoblastic leukemia (B-ALL) with great success.
Methods: To investigate if these therapies alter the immunophenotype of normal B-cell precursors (hematogones, HGs) in bone marrow (BM), we retrospectively reviewed the flow cytometry (FCM) results of 35 post-CART-19 treatment BM specimens from 12 B-ALL patients, 23 post-Blinatumomab treatment BM specimens from 6 B-ALL patients, 13 post-IO treatment BM specimens from 3 B-ALL patients, 11 post-Rituximab treatment BM specimens from 4 B-ALL patients.
Results: Eight CART-19-treated  patients (8/12) showed CD19-negative HGs in BM, which were mostly seen in the first 5 months after treatment and were less than 0.5% of total nucleated cells. Except the absence of CD19, the HGs showed normal maturation pattern.  All the patients (4/4) treated with Rituximab showed CD20-negative HGs with normal expression of other antigens.  One Blinatumomab-treated patient (1/6) showed CD19-negative HGs, which were seen in the first two months after treatment and showed no other phenotypic aberrancy. No patients (0/3) treated with IO showed CD22-negative HGs.
Conclusions: These results have confirmed the presence of phenotypically altered HGs in B-ALL patients treated with CART19, Rituximab and Blinatumomab. Since searching for the difference from normal is the key strategy for FCM MRD detection, the knowledge of these alterations is important to avoid erroneous interpretation of FCM MRD results in patient post immunotherapy.

Best Student Presentation Award Finalist
Integration Of The Ki-67 Proliferation Index With The Ogata Score Improves Its Diagnostic Sensitivity For Myelodysplastic Syndromes
Stefan Mestrum1,2, Eline Cremers3, Norbert de Wit3, Roosmarie Drent2, Frans Ramaekers1,4, Anton Hopman1, Math Leers2
1Department of Molecular Cell Biology, GROW-School for Oncology and Developmental Biology, Maastricht University Medical Center, Maastricht, Netherlands, 2Department of Clinical Chemistry & Hematology, Zuyderland Medical Center, Sittard-Geleen, Netherlands, 3Central Diagnostic Laboratory (CDL), Maastricht University Medical Center, Maastricht, Netherlands, 4Nordic-MUbio, Susteren, Netherlands

Introduction: Technical developments and standardization of flow cytometric analyses in the routine diagnostic work-up of myelodysplastic syndromes (MDS) led to the introduction of the Ogata score, based on four parameters assessing characteristics of blast cells and of myeloid cells. Although the Ogata score has a high specificity (95%) in differentiating between MDS and non-clonal cytopenic controls, its sensitivity is extremely low (44%). This resulted in limited clinical application of the scoring system and, therefore, additional objective parameters are required. The Ki-67 proliferation index is well known for its diagnostic and prognostic applications and routinely implemented in histopathological analyses of solid tumors and non-Hodgkin lymphoma. Therefore, we investigated the diagnostic performance of the Ki-67 proliferation index in bone marrow (BM) cell populations in the differentiation between MDS and cytopenic non-MDS patients.
Methods: BM aspirates of 38 MDS patients and 20 non-clonal cytopenic control patients were analyzed by flow cytometry using specific antibody panels to determine the Ogata score and Ki-67 proliferative indices in BM cell populations.
Results: Ki-67 proliferation indices alone could be used to differentiate between these patient groups as an independent parameter with a sensitivity of up to 76% and specificity of up to 75%. Addition of the Ki-67 proliferation index of blast cells and of myeloid cells to the Ogata score increased its sensitivity from 44% to 92%, while maintaining a high specificity of 86%.
Conclusions: Including the Ki-67 proliferation index of blast cells and that of myeloid cells into the conventional Ogata score significantly improves the flow cytometric diagnosis of MDS. Improvement of the sensitivity of the Ogata score by inclusion of the Ki-67 proliferation indices increases its negative predictive value. This leads to a more reliable classification of cytopenic non-MDS patients, which decreases the number of patients that are falsely diagnosed as pre-leukemic, thereby preventing unnecessary follow-up visits to the clinic.

5:45 - 6:45 PMVirtual Room 4
ORAL ABSTRACTS 4: Molecular/Red Cell/Standard & Quality

Clonal Hematopoiesis In Patients With Cardio-Cerebrovascular Diseases
Jin-Yeong Han, Min-Sun Kwak, Ri-Young Goh, Jae-Ryong Shim, Moo-Hyun Kim
Dong-a University College Of Medicine, Busan, South Korea

Introduction: Clonal hematopoiesis is when hematopoietic stem cell starts making blood cells that have the same changes in their DNA. The blood cells with the genetic mutation are different than the rest of blood cells. One of the surprising risk association is the connection between clonal hematopoiesis and cardiovascular diseases. In this study, we investigated the mutations profiles using the next generation sequencing (NGS) in patients admitted to Busan Regional Cardio-cerebrovascular Center, Busan in Korea.
Methods: NGS analysis was done in the form of targeted sequencing using DNA isolated from the peripheral blood and Illumina MiSeqDx. The panel employed was myelodysplastic syndrome (MDS) panel including 49 prevalent genes. Somatic variants detected were correlated with clinical and laboratory findings.
Results: In acute coronary syndrome patients, tier 3 mutations were found in 66.7% and DNMT3A mutation was one of them. Interestingly, in acute cerebral infarctions, also about 66.7% of patients showed at least tier 3 variants and 50% of them showed tier 1 variants in DNMT3A genes. In control group, there were no tier 1 variants identified.
Conclusions: The most frequently mutated genes in clonal hematopoiesis were reported to be DNMT3A, TET2, and ASXL1. The prevalence was much higher than expected. More thorough understanding of clonal hematopoiesis and associations with hematologic malignancies and cardiovascular diseases is warranted.

Best Student Presentation Award Finalist
Cut-Off Values Of Red Blood Cell Indices In Silent Carrier State And Α-Thalassemia Trait
Dung Nguyen Ngoc, Yen Pham Hai, Thanh Nguyen Ha, Khanh Bach Quoc
National Inst of Hema & Blood Trans, Hanoi, Vietnam

Introduction: Background: α-thalassemia trait and silent carrier state are very common in Vietnam. The rate of non-diagnosis is very high. Because they have no clinical symptoms, hemoglobin electrophoresis is normal and genetic analysis is expensive. Red blood cell indices are important in screening, oriented diagnosis. Purpose: to assess of cut-off values of red blood cell indices in silent carrier state and α-thalassemia trait.
Methods: Retrospective and observational study of 1863 cases (1172 silent carrier state (–α/αα), 215 (–α/–α) thalassemia, 476 (––/αα) thalassemia) who from 15 years old and without iron deficiency in National Institute of Hematology and Blood Transfusion from 1/2017 to 12/2017.
Results: Among silent carrier state and α-thalassemia trait, MCV showed an AUC of 0.947 and the cut-off point of 77.45 provided a sensitivity of 91.6% and a specificity of 88,9%. MCH showed an AUC of 0.973 and the cut-off point of 23.65 provided a sensitivity of 93.6% and a specificity of 92.6%. In α-thalassemia trait, MCV showed an AUC of 0.869 and the cut-off point of 72.15 fl provided a sensitivity of 82.3% and a specificity of 78.4%. MCH showed an AUC of 0.865 and the cut-off point of 21.85 pg provided a sensitivity of 75.8% and a specificity of 81.5%. Thresholds of MCV 78.45 fl and MCH 23.95 pg were capable of highly differentiating between (–α / αα) thalassemia and (–α / –α) thalassemia with sensitivity 88.4% and 91.6% and specificity 75.3% and 85.6%.
Conclusions: Based on the results of genetic analysis of 1863 cases, we found that: Threshold of MCV 77.45 fl and MCH 23.65 pg is capable of distinguishing silent carrier state and α-thalassemia trait with high sensitivity and specificity. Threshold of MCV 72.15 fl and/or MCH 21.85 pg have the ability to suggest (–– / αα) thalassemia.

Hemoglobin A1C Analysis And Hemoglobin Variants
Eleni Petridou, Agapi Kotanidou, Pantelis Mourgos, Ioanna Gavridou, Athanasia Agorasti
Hematology Laboratory,General Hospital of Xanthi, Xanthi, Greece

Introduction: Hemoglobin A1c (HbA1c) reflects the average blood glucose levels during about the preceding trimester and it is used for both the diagnosis and the management of diabetes mellitus. According to the recommendations of The National Academy of Clinical Biochemistry (The Academy of the American Association for Clinical Chemistry), HbA1c results below the lower limit of reference interval should be verified by repeat testing. The aim of our study is to record all sample results below 4.5% NGSP (26 mmol/mol IFCC) during one year.
Methods: The method we use in our laboratory for HbA1c determination is the cation-exchange high performance liquid chromatography (HPLC) on Glycohemoglobin Analyzer HLC-723G8. We retrospectively recorded all HbA1c results below 4.5% NGSP (26 mmol/mol IFCC), the results of the repeated analysis of these samples, the results of the followed HPLC hemoglobin fractionation on β-thalassemia Mode Analyzer HLC-723G7, as well as the interpretation of test results.The paired samples t-test was applied for the comparison of the repeated tests
Results: Out of 6300 HbA1c measurements we obtained fifteen results below 4.5% NGSP (26 mmol/mol IFCC) (Table 1). The repeated analysis of these samples confirmed the good precision of our method (no statistical difference in paired samples t-test). The hemoglobin fraction analysis revealed that all patients had a hemoglobin variant (Table 1). All reports included information about the detected hemoglobin variant and comments on the interferences and the reliability. Table 1. Results of HbA1c analysis & haemoglobinopathy investigation
  HLC-723G8 TOSOH® HLC-723G7 TOSOH®
HbA1c (% NGSP) HbA1c (mmol/mol, IFCC) Hemoglobinopathy investigation
1 2.0 0 homozygous Hb S receiving hydroxyurea
2 2.4 3 beta0-thalassemia / HbS
3 2.2 1 beta0-thalassemia / HbS
4 4.4 25 beta+-thalassemia / HbS
5 0.0 0 homozygous HbO-Arab
6 0.0 0 homozygous HbO-Arab
7 0.0 0 homozygous HbO-Arab
8 3.0 10 homozygous HbO-Arab
9 3.1 10 homozygous HbO-Arab
10 2.7 6 homozygous HbO-Arab
11 2.6 5 homozygous HbO-Arab
12 1.3 0 homozygous HbO-Arab
13 1.9 0 homozygous HbO-Arab
14 0.0 0 homozygous HbO-Arab
15 4.1 21 beta+-thalassemia /  HbO-Arab


Conclusions: As haemoglobinopathies represent an increasing global health issue due to multi-ethnic migration,a hemoglobin variant investigation should be offered in all patients presenting HbA1c results below the lower limit of reference interval. The report should include the necessary information for the interpretation of the examination results.

Using Big Data To Determine Statistically (And Probably Clinically) Important Changes In Ferritin, Iron, Total Iron Binding Capacity And Iron Saturation
Yuelin Qiu3, George Cembrowski1,2, Gwen Clarke4, Daniel Holmes5, Qian Wang6, Christopher McCudden7
1CCQCC/LABORATORY CONCISION, Edmonton, AB, Canada, 2Laboratory Medicine and Pathology, University of Alberta, Edmonton, AB, Canada, 3Faculty of Engineering, University of Alberta, Edmonton, AB, Canada, 4Canadian Blood Services, Edmonton, AB, Canada, 5Providence Health Department of Pathology and Laboratory Medicine, Vancouver, BC, Canada, 6Department of Biochemistry, University of Alberta,, Edmonton, AB, Canada, 7Department of Pathology and Laboratory Medicine, Division of Biochemistry, University of Ottawa,, Ottawa, ON, Canada

Introduction: Biologic variation (BV) and analytical variation (AV) are used to calculate the reference change value (RCV), the minimum difference between 2 results that is statistically significant.  While BV is usually assessed in healthy subjects, it is not clear whether similarly derived BVs are applicable in subjects with abnormal tests.   We devised a unique methodology that generates total variation from serial  patient data (a mixture of preanalytic variation, BV and AV).  Here, we determine statistically significant between day and between week changes in iron-related tests.  
Methods: A data repository provided approximately 346,000 ferritins, 92,000 serum irons, 88,000 total iron binding capacities (TIBC), and 88,000 iron saturations (%), measured over 5 years in inpatients and outpatients at Ottawa Hospital, Canada. For each analyte, we tabulated the pairs of intra-patient results that were separated by 1) 3 hour intervals, 0-3, 3-6, 6-9, 9-12,… up to 48 hours and 2) weekly intervals up to 12 weeks. The standard deviation of duplicates (SDD) of the paired iron analyte determinations was calculated for each time interval. The graphs of SDD vs. time interval were approximately linear; the y intercept provided by the linear regression equation represents the sum of BV and AV: y0=(AV2 + BV2)½.  The longer term variation was assessed from the analysis of 12 contiguous weeks of intrapatient results.  The y intercept was transformed into a percentage by dividing by the patient mean and multiplying by 100.  The between day and between week y intercepts for the four iron tests were multiplied by the factor 21/2 x 1.96 to provide statistically significant changes of p< 0.05.
Results: Means of all of the tests are at the lowest limit of the reference interval.  The RCV for weekly iron and ferritin in patients with deficient iron stores, are 75% and 87%, respectively, indicating that iron must increase by 75% or ferritin by 86% to demonstrate a statistically significant change (P< 0.05).
Conclusions: Use of this SDD-calculated RCV will eventually be incorporated into laboratory decision support software.

5:45 - 7:45 PMVirtual Room 5
VIRTUAL POSTER SESSION

Best Student Presentation Award Finalist
P1
Investigation Of A New Parameter For Enumeration Of Cd34+ Cell In Hematopoietic Stem Cell Identification In Peripheral Blood Stem Cell Harvests Using Sysmex Xn-Hpc &Ndash; A Single Centre Experience
Siew Lian Chong1, Asral Wirda Ahmad Asnawi1, 2, Tengku Amatullah Madeehah Tengku Mohd2, Kim Wah Ho1, Sharifah Shahnaz Syed Abd Kadir1, Veena Selvaratnam1, Sen Mui Tan1
1Ampang Hospital, Kuala Langat, Malaysia, 2Faculty of Medicine and Health Sciences, Universiti Sains Islam Malaysia, Seremban, Malaysia

Introduction: CD34+ cell is used to define cell harvest goals. Successful Peripheral Blood Stem Cell (PBSC) transplantation depends on the infusion of an appropriate number of HPCs to achieve rapid and durable haematologic recovery. Flowcytometry CD34 cell enumeration has been the widely accepted method, however is expensive, time consuming and labour-intensive as it requires  specific equipment and experienced staff. Hence we evaluated the use of the Hematopoietic Progenitor Cell count programme on the Sysmex XN haematology analyser (XN-HPC) as a new parameter for enumeration of CD34+ cell in hematopoietic stem cell.
Methods: One hundred and fifty-one samples were collected from subjects who underwent PBSC collection for allogeneic and autologous stem cell transplantation. Subjects comprising of all ages, both genders and three principal races were recruited. Peripheral blood was used for CD34+ enumeration using FACS Calibur flow cytometry as per standard protocol. The same sample was then used for XN-HPC enumeration. 
Results: Preliminary data was available in 51 patients, comprising of patients who underwent autologous or allogeneic stem cell transplantation and healthy donors. The comparison was then analyzed by Passing Bablok Regression Fit. Kendall Tau's correlations coefficient shows positive correlation at 0.50 (p< 0.001). It also shows linear association with Cusum Test (A=1.027) p =0.24). Results from the Passing Bablok show that the intercept at -2.33 (95% CI -6.798, 0.430) with the slope at 1.18 (95% CI 0.577, 1.590). XN-HPC showered there is statistically significant evidence that the methods differ by at least a constant amount.
Conclusions: XN-HPC showed good analytical performance.With the increasing prevalence of stem cell transplantation and its high costs per year, XN-HPC appears to be suitable, rapid and inexpensive alternative for CD34 cell enumeration.

P2
Could Blood-Based Inflammatory Biomarkers Have A Diagnostic Or A Prognostic Value In Secondary Hemophagocytic Lymphohistiocytosis?
Intissar Helali1,2, Aya Chakroun2, Hela Baccouche2, Sonia Mahjoub2, Neila BenRomdhane2
1Medicine School of Tunis, Tunis, Tunisia, 2Laboratory of Hematology, La Rabta Hospital, Tunis, Tunisia

Introduction: Secondary Hemophagocytic lymphohistiocytosis (sHLH) is a rare but potentially life-threatening systemic infammatory disorder. It complicates infections, cancers, and dysimmune diseases. Its poor prognosis calls for early diagnosis. Simple and easily available, the neutrophil-to-lymphocyte ratio (NLR), the platelet-to-lymphocyte ratio (PLR) and the lymphocyte-to-monocyte ratio (LMR) are gaining attention as diagnostic and prognostic markers for several diseases. In fact, those Blood-Based Inflammatory Biomarkers (BBIB) reflect the systemic inflammatory response. The present study aimed to assess the reliability of NLR, PLR and LMR as predictif tools for sHLH diagnosis and prognosis.
Methods: From June 2017 to January 2019, a prospective and evaluative study collected 53 adult patients suspected as having sHLH referred to a single Hematology Department. The final diagnosis of sHLH was based on the HLH-2004 diagnostic criteria. Two groups were formed: "sHLH group” (n=32) and "not sHLH group” (n=21). Clinical and biological data were collected at baseline and after 6months follow-up. The BBIB were calculated and their correlation with the diagnosis of sHLH and the survival were evaluated. Statistical study was performed using SPSS 23.0 software. Receiver operative characteristic (ROC) curve and Area under the curve (AUC) were determined.
Results: The median age of all patients was 43 years [17-80]. Among clinical and biological criteria, temperature, ferritinemia, triglyceridemia, LDH and H-Score were significantly higher in “sHLH group” (p=0.02, p< 0.01, p< 0.01, p=0.02 and p< 0.01 respectively). No significant statistical differences were shown according to NLR, PLR and LMR values (p=0.743, p=0.785 and p=0,383 respectively) between the two groups. Spearman’s coefficient concluded to no correlation between these ratios and the sHLH diagnosis. Only LMR showed a limited ability (AUC=0.571) to predict the sHLH while NLR and PLR failed to (AUC= 0.473 and 0.478, respectively). In the “sHLH group”, 13 patients died, with a 6-month survival rate of 59%. No correlation was found between NLR, PLR or LMR and the occurrence of death.
Conclusions: In our study, BBIM didn’t show significant effectiveness in diagnosing sHLH, this can be explained by the inflammatory status in both studied groups and the cytopenias characterizing sHLH patients. Likewise, these markers did not show any prognostic value.

P3
Applicability Of Abbott Alinity Hq Hematology Analyzer For Automated Cell Counting Of Body Fluids
Pauline Herroelen, Serge Damiaens, Simke Demeester, Kristin Jochmans
Department of Hematology, Universitair Ziekenhuis Brussel, Brussels, Belgium

Introduction: Body fluid cell count provides essential information for diagnosis and monitoring of diverse pathologies. We evaluated the performance of the novel Alinity HQ hematology analyzer for automated cell counting in body fluids and compared red blood cell (RBC) and white blood cell (WBC) counts with the Cell-Dyn Sapphire automated hematology analyzer.  
Methods: New background concentration limits were defined based on 20 consecutive background runs. Limit of detection (LOD) was derived from 10 replicate measurements of a cell-free cerebrospinal fluid (CSF). Linearity of the analytical measuring range was evaluated by serially diluting a CSF with high RBC and WBC counts. Functional sensitivity, defined as the lowest cell concentration that can be measured with a maximum coefficient of variation (CV%) of 20%, was derived from the imprecision experiment, using a serially diluted pleural fluid. Carryover was evaluated following ICSH guidelines. For method comparison, we compared 172 body fluids, using Passing Bablok analysis, Spearman rank correlation coefficient (rS) and Wilcoxon signed rank analysis.
Results: Background concentration limits were ≤1,000 cells/μL for RBC counts and ≤3 cells/μL for WBC counts. The LOD was 1,000 RBC/µL and 5 WBC/µL. Results from linear regression analysis (table 1) revealed excellent linearity.  CV% was inversely correlated with cell concentrations, ranging from 0.7% to 37.2% for 777 x 103 RBC/µL and 2 x 103 RBC/µL, respectively and from 3.6% to 43.8% for 575 WBC/µL and 5 WBC/µL, respectively. Functional sensitivity was 3,000 cells/µL for RBC counts and 50 cells/µL for WBC counts. Carryover was 0.6% and 0.1% for leukocytes and erythrocytes, respectively. Table 2 displays results of method comparison between Alinity HQ and Cell-Dyn Sapphire. Significant differences were observed for WBC and RBC counts in serous fluids and for RBC counts in CSF. However, Bland Altman analysis revealed acceptable mean differences of less than 20% for all comparisons.
Conclusions: We demonstrated comparable performance in body fluid cell counting between the Alinity HQ hematology analyzer and the Cell-Dyn Sapphire. This makes the Alinity HQ applicable, starting from concentrations of 50 WBC/μL and 3,000 RBC/μL. At least one background run should be performed for carryover avoidance.

P4
Soluble Carcinoembryonic Antigen Cell Adhesion Molecule 1,6 And 8 In Acute Myeloid Leukemia: Their Relation To Survival And Prognosis
Nahla Ibrahim
KAMC

Introduction:  The carcinoembryonic antigen cell adhesion molecules (CEACAM) play important roles in cell adhesion as well as cancer cell invasion and metastasis. Objectives: to study the soluble CEACAM 1,6 and 8 in acute myeloid leukemia (AML) and to determine if they had an impact on the survival and prognosis.
Methods:  Methods: 102 subjects were included. They were 53 with AML and 49 healthy persons. All were subjected to the measurement of soluble CEACAM 1,6 and 8 by ELISA. The patients were divided into the high and low group by using median of each parameter in patients as a cut off value.
Results:  Results. Significant increase of sCEACAM1,6 and 8 was found in their high group when compared to the control group. No significant difference was found in the low group of both sCEACAM1 and 6 when compared to the control. In contrast, a significant increase of sCEACAM8 was found. There were significantly positive and negative correlation of the high sCEACAM1 with lactic dehydrogenase and each of the surface CD66a, sCEACAM6 and 8 respectively. Significant positive correlations were found between sCEACAM6 and 8. There was a Significant increase of the relapse-free survival (RFS) in the highest group of sCEACAM6. Also, it was associated with increased overall survival (OS) 6.2 times when compared to the low group. Soluble CEACAM8 had a significant good impact on induction remission.
Conclusions: Conclusion: sCEACAM1 is a poor prognostic factor. The  high group of sCEACAM6 is a good independent  prognostic factor, whereas sCEACAM8 could predict  a good response to induction remission. High values of sCEACAM6 and sCEACAM8 can suppress or decrease the secretion of sCEACAM1. Recombinant sCEACAM 6 and 8 could be used to counteract the undesired effect of sCEACAM1 and to improve the prognosis of the AML patients. Recommendation The measurement of sCEACAM1 and 6 at diagnosis is essential to identify those with poor prognostic factors. More studies are needed to clarify the role of sCEACAM1, 6 and 8 in AML and to prove our results

P5
Evaluation Of A Prototype Software For Image Analysis Of Reticulocytes
Veronika Jenei, Melina Lundqvist, Mats Erikson, Hans-Inge Bengtsson
CELLAVISION, Lund, Sweden

Introduction: Reticulocytes are immature red blood cells containing residual RNA, ribosomes and mitochondria. Their number in peripheral blood measures the erythropoietic activity of the bone marrow and is important for diagnosing anemias and monitoring bone marrow response to therapy. Although modern cell counters/flow cytometers offer automated reticulocyte counts, manual counting is still used as a reference method and as a cheap alternative despite the relatively high (25-50%) interobserver variation. The aim of our study was to evaluate the usability of image analysis for semiautomated reticulocyte counting.
Methods: A prototype software was developed on the CellaVision® DC-1 digital morphology analyzer. It can analyze over 10 000 red blood cells, count the % of reticulocytes on supravitally-stained smears and present digital images of classified cells. The prototype software was developed using new methylene blue-stained (RAL Diagnostics) blood films using healthy blood or stabilized blood controls (BD Biosciences). Stain stability was monitored over a 9-month period by repeatedly rescanning the same area on the slide and comparing staining intensity. Reticulocyte maturation was tracked in healthy blood at room temperature overnight. The accuracy of the prototype software was tested by comparing pre-classified reticulocyte percentages in stabilized controls with known concentration of reticulocytes to that after re-classification by two independent observers. Repeatability and correlation tests were performed using three-level, stabilized controls.
Results: Staining with new methylene blue at 37 °C for 5-15 minutes gave good results with little water damage or artefacts. Staining intensity was stable during a 9-month period with no or minimal fading observed. Compared to fresh healthy blood, reticulocyte percentage was reduced by 18% and 17% over 24h at room temperature considering pre-classified or re-classified results, respectively. Our results showed an overall good correspondence between pre-classified and re-classified reticulocyte percentages using stabilized controls. There was a linear correlation between pre-classified and re-classified reticulocyte percentages in the range of 1-8.2%, with an R2 value of 0.9872. Reticulocyte results with the prototype software showed high repeatability with a CV of 7.3, 6.6 and 2.5% for low, medium and high-level reticulocyte controls, respectively.
Conclusions: Our results show that digital image analysis can be an acceptable alternative to manual counts. While further development of our prototype software is needed our study showed good correlation with manual counts and good repeatability.

Best Student Presentation Award Finalist
P6
The Added Value Of Laboratory Hematology Data In The Diagnosis Of Ehrlichiosis
Lily Mahapatra, John Frater
Washington University School of Medicine, St. Louis, MO, United States

Introduction: Laboratory diaganosis of ehrlichiosis traditionally relies on strategies that have a long turnaround time, such as polymerase chain reaction (PCR), which may be impractical in some practice environments.  To resolve this limitation, we evaluated the performace of clinical laboratory data from an automated hematology analyzer (Sysmex XN-10) in predicting the diagnosis of ehrlichioisis. 
Methods: We performed a retrospective review of 34 cases of PCR-confirmed ehrlichiosis evaluated using a Sysmex XN-10 analyzer with concurrent complete blood counts and complete metabolic profiles. Two control groups were retrospectively identified. The disease control (DC) group included patients that had a clinical history of tick exposure, but without detection of Ehrlichia by PCR. The healthy control (HC) group included patients who were seen for routine physical exam, without evidence of fever, fatigue, rash, or tick exposure at the time of evaluation. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of detection of ehrlichiosis by abnormal WBC scattergram with and without abnormal blood counts were also calculated.
 
Results: Abnormal scattergrams, demonstrating an increase in atypical or variant lymphocytes, were identified in 11 (32%) PCR-confirmed cases and 5 (14%) of PCR-negative cases (Figure 1). Of the hematologic abnormalities studied, thrombocytopenia emerged as the strongest predictor of ehrlichiosis when comparing the PCR-confirmed cases with the DC or the DC with the HC (Figure 2). The sensitivity and positive predictive value (PPV) for the diagnosis of ehrlichiosis by abnormal WBC scattergram were 32% and 69%, respectively. The sensitivity increased to 94% and the PPV was 62% when thrombocytopenia was included.
Conclusions: Automated hematology coupled with robust evaluation of hematological and chemistry parameters, is a useful and time-effective tool for early diagnosis of ehrlichiosis.
 

P7
Changes In Blood Count In Covid Patients: Hematimetry, Monocyte Distribution Width (Mdw) And High Fluorescence Cells.
Matheus Meira, Rosana Penteado, Claudio Mendes, Andrea Villarinho, Camila Monteiro, Laiz Bento, Priscila Miyamoto, Nydia Bacal, João Guerra
HIAE, Sâo Paulo, Brazil

Introduction: COVID-19 has consolidated itself as a global epidemic, as main etiological agent SARS-CoV-2, infecting millions of people worldwide. Some changes in hematological examinations of diagnosed patients have been described by researchers and scientific groups from several countries. Some of these changes may be useful in suspecting the diagnosis or predicting a more severe course of the disease. This study aimed to identify the main changes in the blood count and investigate specific parameters - hematimetric indices, MDW, IMFI - CD64 and high fluorescence cells.
Methods: We analyzed 100 blood counts of patients with PCR positive for COVID-19, processed in routine analyzers (Sysmex Xn-10® - Blood count, Coulter DxH-900® - MDW parameter, Flow Cytometry – Mean Fluorescence Intensity (MFI) of CD64 in neutrophils). The results obtained were analyzed through nonparametric statistical tests (Mann-Whitney Test, T-student Test, Chi-square Test, Paired T-student Test). A control group with n=24 was used for comparison. The study was based on retrospective observational analyses.
Results: There was a decrease in the count of Red Blood Cells (CBR), Hemoglobin (HB) and Hematocrit (HT) of the COVID-19 group in relation to the control group. Conversely, the count of total leukocytes (WBC), neutrophils, the immature fraction of granulocytes (GI), MDW and MFI of CD64 were higher in patients in the COVID-19 group. Lymphopenia was the most common finding in patients with COVID-19, but also the eosinophil count was lower than the control group. Contrary to the general trend of lymphopenia, high fluorescence lymphocytes were increased in patients with COVID-19, presenting cytoplasmic and nuclear morphological alterations. MFI - CD64 and MDW have diagnostic value as a predictor of Sepsis and presented some sensitivity to the infectious condition caused by the disease, however they cannot be considered specific predictors for COVID-19 and their values should be interpreted with caution.
Conclusions: Based on the results, the parameters investigated in the Blood Count, together with other methodologies of the clinical laboratory, may be useful for a better screening of COVID-19 patients. We conclude that the Clinical Laboratory of Hematology has the important role of providing the clinical team with several useful prognostic markers through the Blood Count, for a more efficient approach in the treatment of these patients.

P8
Application Of The Monoscore To The Differential Diagnosis Between Chronic Myelomonocytic Leukemia And Other Causes Of Monocytosis
Ignacio Molero-Vilches, Alfredo Bermejo, Javier Garralda-Fernandez
Hospital de Fuenlabrada, Madrid, Spain

Introduction: Monocytosis is a common finding on routine blood counts. Chronic myelomonocytic leukemia (CMML) is a myelodysplastic/myeloproliferative neoplasm which may cause monocytosis. Optimal detection of CMML is important from the prognostic and therapeutic point of view. Persistent absolute monocytosis (> 1000 monocytes/μL) is included in the diagnostic criteria for CMML defined by the WHO, along with cytological and molecular findings. A predictive index has been developed for CMML, the Monoscore (Schillinger F, 2018). This index uses basic and structural parameters of the complete blood count (CBC) obtained on the Sysmex XN™ analyzers. A Monoscore value above 0.16 is considered highly predictive of CMML. The aim of this study is not only to determine whether Monoscore is useful in distinguishing CMML from other causes of monocytosis, but also whether it differentiates CMML equally from the various causes of monocytosis.
Methods: We performed a retrospective observational study on 1222 patients with monocytosis (> 1000 monocytes / μL) detected in a CBC performed on a Sysmex XN™ series analyzer. This study was developed in the Laboratory of the Hospital de Fuenlabrada (Madrid, Spain), between October and December 2018, in patients from emergency, hospitalization and outpatient settings. The CBC data used in Monoscore calculation were neutrophil %, monocyte %, absolute monocyte value and neutrophil positional parameter Ne-WX. Once the data had been collected and calculated, the sample was divided into 6 diagnostic groups: Myelodysplastic Syndrome (MDS), neoplasm undergoing chemotherapy, Chronic Myeloproliferative Neoplasm (CMN), Acute Myeloblastic Leukemia (AML), reactive monocytosis and CMML. Monoscore medians of each group were compared with that of CMML, using the U-Mann-Whitney test. ROC curves were also constructed between CMML and each of these groups, and the area under the curve (AUC) was calculated.  For the statistical analysis, MedCalc® version 11 program was used.
Results: Data of 9 MDS, 217 neoplasms, 7 CMN, 5 AML, 10 CMML and 974 reactive monocytosis were analyzed. The comparisons between CMML and the other groups (U-Mann-Whitney) and the AUCs of the ROC curves are presented in Table 1.
Conclusions: In our patients, Monoscore makes it possible to adequately discriminate CMML from other causes of monocytosis, except in the case of AML. The use of Monoscore can guide specific studies aimed at the diagnosis of CMML, thus adequately optimizing laboratory resources.

P9
  The Evaluation Of The Variation Of Rdw-Sd And Rdw-Cv In Covid 19.
Amine Moueden, Reda Messaoudi , Driss Benlaldj, Fatima Seghier
Hemobiology department CHU of Oran , Oran, Algeria

Introduction:  Red cell distribution width (RDW) tells us about the variation in erythrocyte size. An RDW higher than the normal value indicates erythrocyte anisocytosis. In COVID-19, the variation in this parameter is rarely taken into consideration by clinicians, although it is a very important parameter and can provide information in the clinical course and therapeutic management.     .
Methods: The study consists of evaluating the variation of the 02 parameters RDW and PDW in 70 patients with COVID-19. Blood samples are taken on an EDTA (Ethylene Diamine Tetra Acetic) tube. RDW and PDW are directly provided by the Mindray BC - 6800 blood count analyzer. RDW is expressed in 02 forms RDW is expressed in 02 forms RDW-SD (Normal range: 40 % – 55%) and RDW-CV (Normal range: 11%-15%).The neutrophil-lymphocyte ratio (NLR) is an inflammatory biomarker is identified as an independent risk factor for critical illness in patients with COVID-19 infection. The NLR is defined by the absolute number of neutrophils divided by the absolute number of lymphocytes. We compare here the means of RDW-SD and RDW-CV of two groups (NLR ≤ 3, 13 and NLR > 3, 13) and (NLR ≤ 6, 5 and NLR > 6, 5).
Results:
 22 out of 70 patients (31, 4 %) have RDW-CV higherthan 15% (17, 94 ± 3, 32). 08 out of 70 patients (11, 4 %) have RDW-SV higherthan 55 fl (63, 62 ± 6, 32). The mean value of RDW-CV RDW-SD patients with NLR > 3, 13 was quite higher than that of patients with NLR ≤ 3, 13 (13,01± 0,73 /14,90 ±,2,8 p = 0,013 ) ; (40,72 ± 2,05 / 46,45 ± 9,44 p = 0,016 ) respectively but theses values still in normal range, additionally  the mean value of RDW-CV RDW-SD patients with NLR > 6,5 was very higher than that of patients with NLR ≤  6,5 (13,37 ± 1,04 /15,12 ± 3,04 p= 0,009) ;(41,37 ± 3,57/47,26± 9,88 p=0,004).respectively .RDW-CV and RDW-SDcorrelate positively but weakly with NLR (P = 0,011, R = 0, 3); (P = 0,005, R = 0, 33).            
Conclusions: The elevated RDW-CV and RDW-SDare associated with inflammatory marker NLR in COVID 19.Complicating disease could be screened out by increased RDW

P10
Asymptomatic Case Of Covid-19 In A Patient With Acute Myeloid Leukemia. The Role Of Massive Screening.
Naseelah Mussa, Yuliya Volovetska, Fátima Carriço , J. Melo Cristino
Hospital Santa Maria- Centro Hospitalar , Lisbon, Portugal

Introduction: Coronavirus Disease-19 (COVID-19) typically presents with respiratory symptoms and is caused by the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2). The ability of the virus to cause a pandemic situation is related, among others, to the rapidly spread among symptomatic or even people without symptoms (whether truly asymptomatic or pre￾symptomatic), the latter posing a particular risk of massive spread inside heath care facilities. Moreover, multiple co-morbidities can often mask the clinical picture of this infection.
Methods: A 79 year old male with type 2 Diabetes Mellitus, acute myeloid leukemia, and no history of COVID-19 risk factors, was admitted in our emergency room, with a history of fever, a painless nodule in the right axilar region, hematemesis, melena and epistaxis.
Results: Blood tests revealed anemia, leukocytosis with 96% of blastic cells, thrombocytopenia and the nodule revealed to be a right axillary abscess caused by Methicillin Sensitive Staphylococcus aureus￾MSSA. Blood dyscrasia was related to the tyrosine kinase inhibitor Sorafenib. A routine nasopharyngeal swab for SARS-COV-2 RT-PCR was requested prior to the admission in the ward, the result was positive and the patient was shifted to a COVID unit.
Conclusions: People without symptoms readily spread SARS-CoV-2 and may be most infectious before they become symptomatic. Hospital wards triggered measures to minimize the outbreaks mostly by performing a swab for SARS-CoV-2 prior to the admission. In our case the patient never complained about respiratory symptoms, in fact, the patient condition worsened despite the cure for COVID-19 (3 negative swabs), and was transferred to the ICU mostly due to the complications of other comorbidities. However important spreading of the viral infection inside the hospital wards was avoided. Studies have estimated that people with COVID-19 without symptoms could be responsible for up to half of the spread, and minimal or asymptomatic individuals must always be taken in to account.

P11
Analysis Of Mdw As A Biomarker For Early Sepsis Detection In A Tertiary Care Hospital In North India
Shano Naseem1, Neelam Varma1, Ishwar Bihana1, Praveen Sharma1, Brijesh Verma1, Navneet Sharma1, Pallab Ray1, MR Shivaprakash1, RK Ratho1, Sandhya Bastian2, Elena Sukhacheva3
1Post-Graduate Institute of Medical Education & Research, Chandigarh, India, 2Beckman Coulter India Pvt. Ltd., Mumbai, India, 3Beckman Coulter Eurocenter, Nyon, Switzerland

Introduction: Early detection of patients having or developing sepsis in the emergency department (ED) is important for timely administration of specific treatment. Lack of early sepsis recognition is a current challenge that could delay or cause a misdiagnosis and potentially contribute to adverse outcomes. Therefore, identification of an early reliable biomarker of sepsis will be of utmost importance. We studied Monocyte Distribution Width (MDW),a new biomarker available on the hematology analyser DxH 900,as a potential biomarker for sepsis detection in the ED.
Methods: Study included 356 adult patients admitted to ED in a tertiary care hospital in North India.  CBC-Diff analysis was performed at presentation on Beckman Coulter DxH 900, within 2 hours from blood draw, with blood collected in K2-EDTA tubes.Patients were classified into diagnostic groups with sepsis-2 criteria.  The diagnostic performance of MDW was evaluated using MedCalc, v.19.6.4 (MedCalc Software bvba, Ostend, Belgium), calculating the area under the curve (AUC), Sensitivity, Specificity, Positive Predictive Value (PPV) and Negative Predictive Value (NPV).
Results: Patients' groups are presented in Table 1. Median MDW values increased from no-infection/no-SIRS group (21.2), to SIRS (23.3), Sepsis (26.4), Severe sepsis (28.2) and Septic shock (29.3). The results of MDW and WBC were reviewed against patients’ diagnosis and classified in the Table 2, showing all septic patients (including severe sepsis and septic shock) versus all other groups. When MDW was combined with WBC, the sensitivity and NPV reached 100% for sepsis detection in our cohort, if at least one biomarker (MDW or WBC) was abnormal, suggesting the use as an indicator of potential sepsis. For low risk patients (presenting with 1 or 2 SIRS), abnormal MDW results increased sepsis probability 4.7 times compared to normal MDW results (47% vs 10%). However, in the whole cohort, abnormal MDW results increased sepsis probability to almost 7 times (29.2% vs 4.1%) compared to normal MDW result.
Conclusions: MDW demonstrated high sensitivity and high NPV for sepsis detection in the ED patients.   MDW can be considered as a useful tool to aid in early sepsis detection in the ED.

P12
Differential Morphometric Parameters In Patients Infected With Sars-Cov-2 And Influenza A Virus.
Javier Nieto-Moragas1,2, Anna Marull1,2, Jose Manuel Ramirez1, Orlando Jimenez1,2, Francisco Reina2, Raul Benitez3, Maite Serrando1,2
1Laboratori Clínic Territorial de Girona. Parc Hospitalari Martí i Julia. , Girona, Spain, 2Departamento de Ciencias Médicas. Grupo de Investigación en Anatomía Clínica, Embriologíay Neurociencia. Universitat de Girona., Girona, Spain, 3Centre de Recerca en Enginyeria Biomèdica, Escola d’Enginyeria de Barcelona Est, Universidad Politécnica de Catalunya., Barcelona, Spain

Introduction: Laboratory parameters such as lymphocyte count, lactate dehydrogenase, interleukin-6 or D-dimer have provided prognostic information for patients with COVID-19. Cell population data (CDPs) are morphometric data used in hemacytometry that report the characteristics of cells present in peripheral blood. The aim of the study is to determine the usefulness of hemacytometric parameters to differentiate the morphological characteristics of leukocyte subpopulations in patients infected with SARS-CoV-2 versus those infected with influenza A virus.  
Methods: Patients with SARS-CoV-2 or influenza A virus infection detected by polymerase chain reaction (PCR) were included in the study. Complete blood count and 18 CPDs results obtained from the Sysmex XN10 and coagulation profile and C-reactive protein (CRP) at the time of diagnosis were collected for both groups. Group means were compared and a support vector machine with radial basis function (SVM-RBF) was trained to predict the classification of infected patients and validated by 5 fold cross-validation technique.
Results: One hundred thirty-one patients were included and were distributed in the COVID group (n=71) and the influenza group (n=60). Different laboratory results were compared by non-parametric univariate tests methods (Table 1).  
Parameter COVID group Influenza A group P-value
C reactive protein (mg/dl) 10.52 8.08 0.26
Prothrombin time (ratio) 1.21 1.22 1.0
White blood cells (103/µL) 7.18 8.01 0.34
Lymphocytes (%) 13.0 16.1 0.13
Table 1. Results between COVID group and Influenza group.   No significant differences were observed when comparing the means of laboratory results between both groups. An area under the curve (AUC-ROC) of 0.91 (95% CI: 0.89-0.93) was obtained with a sensitivity of 94% and a specificity of 87% to differentiate cases from the COVID group. Using the recursive feature elimination technique, the parameters with the greatest potential to differentiate cases were LY-X and MO-X.
Conclusions: Cell population data obtained from hematology analyzers can detect changes in cell morphology. CPDs can be integrated into a multivariate algorithm to differentiate patients infected with SARS-CoV2 or Influena A virus. Further studies are needed to validate the results and to consider other causes of infection.

P13
Performance Of Monocyte Distribution Width As An Aid In Sepsis Detection In Adult Patients With Monocytopenia And Leukopenia
Chan-Jeoung Park1, A la Woo2, Dong Kyu Oh2, Sang-Bum Hong2, Elena Sukhacheva3
1Department of Laboratory Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea, 2Department of Pulmonary and Critical Care Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea, 3Beckman Coulter Eurocenter, Nyon, Switzerland

Introduction: Early sepsis detection in the emergency department (ED) is critical, but often is delayed due to time needed for clinical assessment and laboratory tests to be ordered and results become available. We recently evaluated new sepsis biomarker, Monocyte Distribution Width (MDW) as an aid in sepsis detection for general adult ED population. As some patients with sepsis present with low WBC count and/or low monocyte count, we were specifically interested to asses the performance of MDW in patients with leukopenia or monocytopenia.
Methods: From 549 patients enrolled in MDW study (presented as abstract at ISLH 2020) there were 73 patients with leukopenia (WBC< 4000 cells/ml) and 97 patients with monocytopenia (absolute monocyte count< 300 cells/ml). Blood samples were collected in K2-EDTA tubes and tested on Beckman Coulter DxH 900 hematology analyzer to obtain MDW results as part of CBC-Diff. Sepsis diagnosis was established with sepsis-3 criteria. The diagnostic performance of MDW was evaluated by ROC analysis, calculating area under the curve (AUC), sensitivity, specificity, Positive Predictive Value (PPV) and Negative Predictive Value (NPV). Statistical analysis was performed with Medcalc software, version 19.6.4 (MedCalc Software bvba, Ostend, Belgium).
Results: 97 patients with monocytopenia and 73 patients with leucopenia were classified into 4 groups according to sepsis-3 criteria, as presented in the Table 1. In the group with monocytopenia there were 57 patients with cancer (58.8%), 15 of them with hematological malignancy.  Among leukopenic patients there were 58 patients with cancer (79.5%), including 12 patients with hematological malignancies.  Performance characteristics of MDW in sub-groups of ED patients with monocytopenia or leukopenia, presented in the Table 2, were like those obtained for the whole cohort: AUC 0.71, sensitivity 81.4% and specificity 50.7% at a cut-off of 20.
Conclusions: Our results demonstrated that performance of MDW as an aid for sepsis detection in the ED is not affected in patients with monocytopenia and leukopenia, even in the cohort with high prevalence of patients with cancer.

P14
Spurious Eosinophilia On Sysmex Xn-1000 Automated Hematology Analyzer
Eleni Petridou, Agapi Kotanidou, Paraskevi Aslanoglou, Maria Theocharidou, Ioanna Gavridou, Athanasia Agorasti
Hematology Laboratory, General Hospital of Xanthi, Xanthi, Greece

Introduction: On Sysmex XN-1000 automated hematology analyzer, the measurement principle for White Blood Cells (WBC) and Differential (DIFF) analysis is fluorescence flow cytometry. The fluorescence marker labels the intracellular DNA and RNA. In the generated scattergrams, the cells are differentiated according to their fluorescence signal and their internal structure. On DIFF scattergram, the neutrophils plus basophils (NEUT + BASO) populations form a turquoise cloud, while the eosinophils (EO) population a red one (Figure 1). The aim of this study is to present ten cases with abnormal shape in the NEUT + BASO and EO clusters. More precisely, the EO cloud overlaps the lower part of NEUT + BASO cloud (Figure 2). 
Methods: Ten cases with abnormal shape in the NEUT + BASO and EO clusters, flagged “Eosinophilia” and equal number of cases presenting eosinophilia with normal clusters were included in the study. We retrospectively recorded the WBC & Differential analysis, the labeling on data and the WBC Flag(s) retrieved by the files data of Sysmex XN-1000 analyzer, as well as the results of the peripheral blood smears review stored in the laboratory information system. In addition, we recorded two parameters indicative for neutrophils activation; NE-SSC representative of the structure and NE-SFL related to production or release of proteins and reactive oxygen representatives.
Results: In all cases with abnormal shape in NEUT + BASO and EO clusters, the results of all the study parameters (WBC, NEUT#, EO#, NE-SSC, NE-SFL) were labeled with an asterisk (*) and all cases were flagged with “WBC Abnormal scattergram” and “Eosinophilia”. The slides review reveals that the eosinophils absolute count was inaccurately measured by the analyzer and the neutrophils presented either several vacuoles and / or toxic granulation. In all cases with normal NEUT + BASO and EO clusters, none of the study parameters were labeled with an asterisk and all cases were flagged with “Eosinophilia”. The study of the peripheral blood smears confirmed the accuracy of the analyzer’s measurement 
Conclusions: When the EO cloud overlaps the NEUT + BASO cloud, the reporting “Eosinophilia” is spurious and the slide review is mandatory. The study of the scattergrams generated by Sysmex XN-1000 hematology analyzer during the full blood count analysis is essential in order to prevent false reports.

P15
  Bone Marrow Examination In Geriatric Patients: About 65 Cases    
riahi salma1, jawedi sahar2, bouattay amina1,2, kortas mondher2
1university hospital sahloul sousse , sousse, Tunisia, 2university hospital farhat hached, sousse, Tunisia, 3university hospital sahloul sousse , sousse, Tunisia, 4university hospital farhat hached, sousse, Tunisia

Introduction: The proportion of elderly is increasing worldwide, especially in industrialized countries, reflecting the aging of the population. Bone marrow examination which is an important diagnostic tool for various diseases may vary in geriatric population where cytopenias are frequently found. The objective is to study the indications of the bone marrow examination in geriatric population and its contribution to the diagnosis.
Methods: A retrospective study conducted in the laboratory of hematology of the University Hospital Farhat Hached of Sousse-Tunisia. This study involved all patients over 65 years with cytopenia, for whom a bone marrow examination was requested, during the period from January 2017 to September 2018.
Results: The median age of our patients is 72.42±0.731 years with a sex ratio of 1.096. We have found that 59 patients had anemia. Thrombocytopenia was reported in 33 patients. Bicytopenia was found in 13 patients and pancytopenia in 17 cases. The bone marrow was inconclusive in 14% of cases, normal in 18 % of cases and in 68% of cases a diagnosis was mentioned. Myelodysplastic sundrome was the most common etiology (25%) followed by multiple myeloma (13.64%), acute leukemias (11.37%)and vitamin deficiency (11.37%). The bone marrow was compatible with a marrow reaction related to an infection or an autoimmune pathology in 11.37% of the cases. Other rare etiologies were found: hairy cell leukemia, medullary metastasis, bone marrow aplasia, iron deficiency.
Conclusions: : Cytopenias in the elderly are quite common and require the implementation of a rigorous diagnostic strategy. Thorough examination of the blood and bone marrow smear is necessary to determine the etiology and guide the therapeutic management.

P16
Analysis Of Mdw* As A Biomarker Of Severity And Outcome In Covid-19 Patients.
Nahid Anis Shaikh1, Rania Medhat Seleim1, Firas Jaafar Kareem Al-Najjar1, Laila Al Suwaidi2, Aaron Han2, Elena Sukhacheva3
1Rashid Hospital Dubai Health Authority , Dubai, United Arab Emirates, 2Mohammed Bin Rashid University of Medicine , Dubai, United Arab Emirates, 3Beckman Coulter Eurocenter, Nyon, Switzerland

Introduction: The novel severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) pandemic has infected 115.6 million people, and caused over 2.6 million deaths worldwide as of March 2021. We studied a novel biomarker, Monocyte Distribution Width (MDW), in COVID-19 patients, evaluating potential utility of MDW as biomarker of disease severity and prognosis in SARS-CoV-2 infection. Monocytes are critical immune cells, and we hypothesized that morphological changes in monocytes, detected by MDW, may correlate with disease severity and assist in predicting patient outcomes. We also analyzed changes in Ly count (LY#), Neutrophil count (NE#) and Neutrophil-to-Lymphocyte-Ratio (NLR) in COVID-19 patients.
Methods: The study included 99 uninfected adult controls and 715 adult COVID-19 patients admitted to ED in a tertiary care facility in Dubai, UAE, during the initial wave of pandemic in 2020. The SARS-CoV-2 RNA was detected by nasal swab reverse transcription polymerase chain reaction.  Blood, collected in K2-EDTA, was analyzed on Beckman Coulter DxH 900, within 2 hours from blood draw. Patients were classified according to disease severity in the following groups: asymptomatic, mild, moderate and severe. MedCalc software v.19.6.4 (MedCalc Software bvba, Ostend, Belgium) was used for statistical analysis.
Results: The COVID-19 patient groups studied included 100 asymptomatic, 107 mild, 189 moderate and 319 severe cases (37 deceased).  Results for MDW, LY#, NE# and NLR, according to severity, are presented in Table 1. We observed increasing trends for MDW, NE# and NLR, and decreasing trend for LY#, moving from the controls to the infected sub-groups. ROC analysis for MDW to predict patient outcomes, presented in Table 2, demonstrated the potential of this novel biomarker as indicator of patient outcomes with specificity of 95.2% at cut-off 30.6.
Conclusions: MDW may be considered as potentially useful and inexpensive tool to aid the management of COVID-19 patients in the emergency department. *For scientific discussion only. The measurement of MDW on the UniCel DxH 900 analyzer is intended for use with adult patients presenting to the emergency department, on whom a white cell differential test has been ordered, as an aid in the early detection of patients with or developing sepsis.

P17
Evaluation Of Atypical Lymphocyte Flag By Sysmex Analyzer On The Basis Of &Ldquo;Q Flag&Rdquo;
Seema Sharma, Vikrant Verma, Riti Yadav, Raghavendra Lingaiah
SGPGIMS, Lucknow, India

Introduction: Automated hematology analyzers generate white blood cell (WBC) suspect flags to indicate smears requiring manual review. One common indication for a manual smear review is the detection of atypical lymphocyte by hematology analyzers. We evaluated the frequency and efficiency of this flag in  Sysmex XE Alpha N  hematology analyzer in our patient population  .
Methods: 12817 consecutive blood samples were reviewed to determine the frequency of atypical lymphocytes in various patient populations in our tertiary care hospital. The atypical lymphocyte flag was generated in 399 samples. The flag for ‘atypical lympho?’ is generated by the WDF channel at the factory setting (Q = 100). We evaluated automated atypical lymphocyte flag on corresponding   peripheral blood (PB) samples with the manual differential count. A total of 200 cells were counted by 2 pathologists each counting 100 cells. The presence of single cells (1 of 200; 0.5%) associated with flag (‘‘atypical lympho’’) qualified as a true-positive result for a blood smear. Enlarged lymphocytes with irregular monocyte-like nucleus and abundant bluish cytoplasm were counted as atypical lymphocytes.
Results: The frequency of atypical lymphocyte flag was 3.1%. The sensitivity of this flag was 98% with a false positivity of 23%. In flagged samples 0.5-1% atypical lymphocytes were present in 31 % while  >5 % were present in 8% in cases. 33.3 % of flagged cases had bacterial or viral infection while in 54.6% cases no definite cause could be ascertained. Q values ranged from 100 to 300 with a mean & SD of 238.5+68.1. The positive predictive value at Q value of 100 was 73% and it increased to 83% at Q value of 150.
Conclusions: The automated method of atypical lymphocyte flagging based on Q values is very sensitive and adjustments in the Q value threshold may help in improving the positive predictive value.
 

P18
Differential Blood Count Changes In Patients With B-Cell Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma (B-Cll/Sll)During Covid-19 Infection
Jasna Stanfel, Ksenija Paradinovic, Blazenka Dobrosevic, Marija Milic, Ana Pocrnja
Osijek University Hospital, croatia, Osijek, Croatia

Introduction: The neutrophil-to-lymphocyte ratio (NLR) increase and presence of lymphopenia are diagnostically significant for assessing the development of severe forms of Covid-19 infections. B-CLL/SLL is a malignant slow-progressing disease with lymphocytosis and a high proportion of non-functional B lymphocytes in the peripheral blood. The early stage of the disease does not require specific therapy. Due to lymphocytosis, NLR is significantly different from healthy individuals. In these patients COVID -19 infections are expected to be critical. The goal is to present changes in differential blood count and NLR for two B-CLL/SLL patients with a mild form of the COVID-19 infectionand moderate increase of CRP.
Methods: Both patients, female (51) and male (60) with not treated B-CLL/SLL disease and confirmed SARS-CoV-2 using RT PCR test have developed only mild symptoms. A complete blood count (Sysmex XN1000) and CRP (immunoturbidimetry, Beckman Coulter AU680)were measured prior and during Covid-19 infection. 
Results: Pre infection values of WBC, absolute neutrophil count (ANC), absolute lymphocyte count (ALC), NLR and CRP were compared with results during infection.NLR and differences between results were calculated. The results are shown in tables.  During COVID-19 infection, WBC count and ALC decreased while NLR increased in both patients. Differences were: WBC= -42.7%, ANC= 15.6%, ALC= -39.2%; Male (60): WBC= -25.2%, ANC= 11.9%, ALC= -35.4%.
Conclusions: Existing studies have already questioned the necessity of B cells involvement in mounting a successful response to SARS-CoV-2 infection. Although it can be expected that patients with B-CLL/SLL could develop a more severe form of the disease due to the presence of B lymphocytes with reduced functionality, they developed only a mild form of the disease. In addition, both patients had a significant decrease in WBC count (42.7 and 25.2%)which is related to the lymphocyte population, while the ANC increased by 15.6% and 11.9%. This has resulted in an increase in NLR, which is still significantly below the NLR for the healthy population. Immunophenotyping could provide an answer to which lymphocyte population the decline is related to. Calculating NLR in a larger number of patients with B-CLL/SLL in COVID 19 infection could provide the stratification risk for these patients.

Best Student Presentation Award Finalist
P19
The Use Of Nlr And Plr As Biological Markers To Determine The Covid-19 Disease Severity Among Patients With Hematological Malignancies.
Victor Tomacinschii1,2, Maria Robu1, Sanda Buruiana1, Aliona Golub2, Cristina Dudnic1,2, Veronica Finciuc1,2
1Nicolae Testimitanu SUMPh, Chisinau, Moldova, 2Department of Hematology, Institute of Oncology of Republic of Moldova, Chisinau, Moldova

Introduction: Patients with hematological malignancies(HM) are more susceptible to and have higher mortality rates from COVID-19 than the general population. As the unfolding COVID-19 pandemic continues to affect the global healthcare systems. Description of affordable but meanwhile precise markers into the determination of the disease severity. The aim of this study is the assessment of neutrophil-to-lymphocyte ratio(NLR) and platelet-to-lymphocyte ratio (PLR) as a potentials biomarker of the severity and morbidity of COVID-19 infection among patients with HM.
Methods: A retrospective study was conducted on 56 COVID-19 positive patients who have been hospitalized between July 2020 and January 2021. Medical records of the cases are analyzed and compared retrospectively for demographic, clinical and laboratory data. The standard deviation, receiver operating curve (ROC) and area under the curve(AUC) were calculated to compare the diagnostic performance NLR and PLR into the determination of the disease severity
Results: From of 56 examined patients, 29(51,2%) were females, the median age of infection was 57 years (21–84 years). The most frequent hematological malignancies among the studies group were: Non-Hodgkin lymphomas(46,2%), followed by multiple myeloma(16%), acute leukemias(10,7%), chronic lymphocytic leukemia(7,1%), and others. 35 out of 56 patients developed a severe form of COVID-19 infection(62,5%), out of them, 27(77,1%) necessitating access into an ICU unit. Medium severity was detected in 16 patients(28,6%), mild form being diagnosed in just 5 of them(8,9%). The mortality rate in the study group constitutes 35,7%(20 of 56 patients). On receiver operating characteristic curve(ROC) analysis, the predictive ability of NLR to detect critical patients was significant (p= 0.003; AUC:0.729), with an optimal cut-off value of 8.375 having 82% sensitivity and 66.7% specificity to determine the severe COVID-19 cases among hematological patients. On the other hand, the ROC curve for PLR has shown an AUC=0,529, p=0.71, statistically insignificant.(Fig. 1)
Conclusions: Severe forms of COVID-19 were diagnosed in 62,5% of cases. The mortality rate in the study group constitutes 35,7%. NLR can be a useful tool in the prediction of severe cases of COVID-19 infection with quite high sensitivity (82%) but with a moderate specificity(66,7%). PLR did not confirm its applicability in the determination of severe COVID cases among HM patients having an AUC=0529, p=0,71.

P20
Reference Change Values Of Leukocyte Morphological Parameters (Cpd) From Mindray Bc8600 Plus Analyzer
ELOISA URRECHAGA1, Mónica Fernández2, Mónica Merino1, Garazi Mugertza1, Cristina Ponga1
1HOSPITAL GALDAKAO USANSOLO, GALDAKAO VIZCAYA, Spain, 2Hospital Universitario Araba, Vitoria, Spain

Introduction: Cell population data (CPD) are reported as part of leukocyte differentials by Mindray BC 6800Plus analyzer, giving information of the size (NE Z, LY Z, MO Z), nucleic acid content (NE Y, LY Y, MO Y) and internal structure (NE X, LY X , MO X) of the  leukocyte subsets.
CPD values reflect in numbers the functional and morphological changes triggered by diverse clinical conditions. Several recent publications have demonstrated the potential clinical applications CPD in various pathological conditions, diagnosing myelodysplastic syndrome, 0preliminary screening of B cell lymphoproliferative disease or acute leukemia,  and early detection of sepsis. The RCV or critical difference is defined as the percentage change that should be exceeded given the analytical and biological variations inherent to a particular test, in that there is a significant difference between the two consecutive measurements. We reported the RCV for CPD reported by Mindray BC8600 Plus analyzer.
Methods: Analytical variation (imprecision, CVa) was assessed using XN CHECK Control  following CLSI EP05, running all control  in triplicate twice each day over five  days. Published biological variation data (CVi) were applied to calculate RCV, according to the equation: RCV =2.77* (CVa2 + CVi2)1/2 for two-sided approach probability 95% P < 0 .05
Results:
  Mean SD CVa CVi RCV
NE-X 393,3 5,26 1,3 0,5 3,9
NE-Y 701,7 12,76 1,7 0,9 5,3
NE-Z 2790,8 12,5 0,4 0,9 2,7
LY-X 75,9 1,03 1,4 0,4 4,0
LY-Y 657,9 9,84 1,4 1 4,8
LY-Z 1480,9 18,63 1,3 0,4 3,8
MO-X 185,8 5,6 3 0,2 8,3
MO-Y 1144,5 28,57 2,4 0,2 6,7
MO-Z 2342,4 41,1 1,8 1 5,7

Conclusions: Documentation of RCV is an essential pre-requisite before the inclusion of new parameters in the laboratory set of reportable tests of general use. The RCV refers to the statistical probability that a change in an individual’s serial results is significant, it is a robust criterion to generate objective delta-check values for use in quality management. Improvement of the delta check process by the laboratory information systems   should be of great interest, generating   flags of significant changes in serial laboratory results from individuals.    

P21
Leukocyte Differential And Morphometric Parameters  From Mindray Bc6800 Plus Analyzer In Rapid Discrimination Of Sars Cov2 Infection
ELOISA URRECHAGA1, Mónica Fernández2, Mónica Merino1, Garazi Mugertza1, Cristina Ponga1, Urko Aguirre3
1HOSPITAL GALDAKAO USANSOLO, GALDAKAO VIZCAYA, Spain, 2Hospital Universitario Araba, Vitoria, Spain, 3Research Unit OSI Barrualde, Galdakao, Spain

Introduction: Morphometric parameters  (CPD) are reported as part of leukocyte differentials by Mindray BC 6800Plus analyzer, give information of the size (NE Z, LY Z, MO Z), nucleic acid content (NE Y, LY Y, MO Y) and internal structure (NE X, LY X , MO X) of leukocytes.  We evaluate leukocyte differential, neutrophil/lymphocyte ratio (NLR) and CPD in the recognition of SARS CoV-2 infection, among others of different etiologies.
Methods: The study group included 65 COVID19 and 105 patients with other infections (30 viral, 75 bacterial) at admission to the Emergency Department.
The diagnostic performance of individual parameters in the differential diagnosis of diverse etiology infections was evaluated using receiver operating characteristic (ROC) curve analysis.  CBC and CPD data were scaled and normalized to z-scores and  unsupervised K-means clustering method was applied, to obtain the optimal number of clusters. The reliability of the model was evaluated in a validation group of 165 patients (89 COVID19, 76 non-COVID19), the percentages of correct classification were calculated.
Results: NLR to  distinguish viral infections from bacterial  and SARS CoV-2 infections, both characterized by neutrophilia:   area under the curve (AUC)  0.980 (95% CI, 0.970-0.991)  cut-off ≤3.31, sensitivity  96.4 % and specificity 93.5 %. The clustering method performed well  to discriminate COVID 19 patients: 70 in 89 COVID19 patients (78.9 %) were correctly classified.
Validation group SARS-CoV-2   Other Viral  infections    Bacterial infections   
neutrophils   109 /L 4.46 3.81 15.4
lymphocytes  109 /L 1.07 2.2 1.56
NLR 4.2 1.3 10.4
NEUT X 361 365 343
NEUT Y 436 440 457
NEUT Z 1847 1934 1857
LY X 102 108 102
LY Y 648 646 664
LY Z 1059 1116 1070
MO X 212 221 209
MO Y 943 950 949
MO Z 1448 1360 1342

Conclusions: NLR and CPD values reflect numerically the functional and morphological changes triggered by infections. Patients could be distinguished into distinct groups based on the etiology of the infection evaluating analytical data altogether.

P22
Serendipitous Detection Of Hb Variants On The  Leukocyte Differential Channel Wdf Of Sysmex Xn Analyzers
ELOISA URRECHAGA1, Mónica Fernández2, Mónica Merino1, Garazi Mugertza1, Cristina Ponga1, Rafael Andres Del Orbe3
1HOSPITAL GALDAKAO USANSOLO, GALDAKAO VIZCAYA, Spain, 2Hospital Universitario Araba, Vitoria, Spain, 3Hospital Universitario Cruces, Baracaldo, Spain

Introduction: Mild hemolytic anemia can be the only clinical evidence of unstable Hb, but minimal changes in CBC don´t trigger flags, so the condition cannot be detected in most cases. The surfactant  in leukocyte   channel in  Sysmex XN analyzer induces the lysis of  red blood cells , unstable Hb variants induce resistance producing low fluorescence signals in the leukocyte differential scattergram (WDF) is detected. Cell population data (CPD) are reported as part of leukocyte differentials, give information of the size (NEUT Z, LY Z, MO Z), nucleic acid content (NEUT Y, LY Y, MO Y) and internal structure (NEUT X, LY X , MO X) of leukocytes. We explored  CPD values in carriers of unstable Hb, to evaluate the interest of including these parameters in the expert rules to generate specific flags to allow the detection of those patients.
Methods: Four families carriers of Hb Shelby, Hb Debrousse, Hb G Philadelphia and Hb G Ferrara have been diagnosed with 7 affected patients.  We study 38 CBC requested in the course of their routine controls. The distinctive WDF scattergram patterns were similar for all cases, revealing low fluorescence signals but no alarm flags were generated. CPD were compared with the values in a group of 200 healthy subjects; differences in mean values were assessed with t Student test , considering significant P < 0.05.
Results: CPD mean values were as follows
  Shelby Debrousse G Philadelphia G Ferrara Healthy P  
NE X 153 155 154 155 148 0.001  
NE Y 36 38 39 38 49 < 0.001
NE Z 90 94 88 88 87 0.01
LY X 85 86 81 81 78 0.01  
LY Y 41 46 47 44 71 < 0.001  
LY Z 61 64 61 59 58 0.01  
MO X 119 120 118 116 117 0.01  
MO  Y 69 80 78 75 114 < 0.001  
MOZ 67 72 68 69 69 0.04  

Conclusions: CPD are distinctively lower in unstable Hb carriers could useful to optimize the detection of samples presenting RBC lysis resistance; their values could  trigger alarms useful to select samples for  further investigation for the presence of hemoglobinopathy.

P23
Reference Intervals Of Neutrophil-To-Lymphocyte Ratio, Platelet-To-Lymphocyte Ratio, Lymphocyte-To-Monocyte Ratio, And Systemic Immune Inflammation Index In Healthy Adults
ELOISA URRECHAGA1, Mónica Fernández2, Mónica Merino1, Garazi Mugertza1, Cristina Ponga1
1HOSPITAL GALDAKAO USANSOLO, GALDAKAO VIZCAYA, Spain, 2Hospital Universitario Araba, Vitoria, Spain

Introduction: Numerous studies have shown that the hematological components of the systemic inflammatory response, including the neutrophil-to-lymphocyte ratio (NLR), the platelet-to-lymphocyte ratio (PLR), the lymphocyte-to-monocyte ratio (LMR), and the systemic immune inflammation index (SII) are efficient prognostic indicators  in patients with a variety of diseases cancer ,  cardiovascular disease and recently SARS Cov2 infection. Our objective was to obtain their reference values in the general population in our area , Vizcaya province  in North Spain.
Methods: A retrospective  study was performed on healthy adult  Caucasian population 18 -80 years  by retrieving the data of routine health controls from our Hospital database and LIS. CBC were analyzed with XN Sysmex 9000 analyze; XbarM control program, routinely  employed to assess the moving averages as a  long-term and continuous control process, was used to select patients, with all CBC parameters within reference ranges and no morphology alarms were selected.  Clinical and Laboratory Standards Institute (CLSI)  document C28-A2 was used to establish reference ranges; outliers were detected using Dixon test. Differences between gender and  age groups (18-45,46-60, 61-80 years), were assessed using the Student’s t-test or ANOVA. To verify if the reference interval should be divided for each group gender /age the Lahti criteria was applied.
Results: 5963 were included in the study (56 % female, 44 % male),  mean age 59.5 years (standard deviation 10.3 years).  Regarding age the population was divided in 2 groups 71 % of them  in the interval  18-60 years (71 %) and 61-80 (29%).
  total male female 18-60 years 61-80 years P
NLR 1.77 0.85-3.9 1.88 0.89-4.0 1.70 0.81-3.7 1.76 0.88-3.92 1.87 0.86-4.04 < .001
PLR 118 60-219 110 57-211 122 65-226 118 49-198 125 61-268 < .001
LMR 5.77 2.95-11.2 5.43 2.93-11.5 5.87 2.99-12.5 5.32 2.69-9.99 5.99 2.99-10.9 < .001
SII   (*109/L) 435 170-1033 431 170-1030 443 184-1039 425 165-1022 459 183-1111   < .001

Conclusions:  Reliable reference values is a requisite to  standardize clinical applications and promote the use of these indicators into the routine complete blood count report.

P24
Monocyte Distribution Width (Mdw): Novel Biomarker Marker Of Disease Severity In Patients With Covid-19 Infection
Neelam Varma, Shano Naseem, Praveen Sharma, Ishwar Chand, Niranjan Khaire, Pankaj Malhotra, Brijesh Verma
PGIMER, Chandigarh, India

Introduction: In Covid-19 infection, around 15–30% patients develop severe or critical disease with a fatality rate of 1–15%, which further increases with age and associated co-morbidities. Thus, identification of patients who may have severe or critical illness is important. We studied the utility of Monocyte Distribution Width (MDW)as potential biomarker for severity detection in Covid-19 infection.
Methods: Prospectively 145 patients with confirmed Covid-19 infection and in whom complete blood count (CBC) with differential leucocyte count (DLC) were performed at presentation, were included in the study. CBC and DLC were done on Beckman Coulter DxH 900. Patients were categorized into 2 groups: Group-1 (n=122) - including asymptomatic (n=57), mild (n=47), and moderate (n=18) patients and Group 2 - including severe patients (n=23). The performance of MDW was evaluated by calculating the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, Positive Predictive Value (PPV) and Negative Predictive Value (NPV). 
Results: MDW at cut-off >18.31 demonstrated 100% sensitivity and 100% NPV for severity prediction. Also, multiple other hematological parameters showed utility for severity prediction- i) NE% at cut-off >49.7%, demonstrated 100% sensitivity and 100% NPV, (ii) NE abs at cut-off >2.4x109/L, demonstrated 100% sensitivity and 100% NPV, (iii) Ly% 25.4 showed 100% sensitivity and 100% NPV for poor outcome.
Conclusions: MDW was found to be useful in predicting severity of disease and outcome. It showed 100% sensitivity and 100% NPV at cut-off >18.31 for severity prediction and at cut-off >25.4 for poor outcome. As CBC is a routine test ordered in every patient at time of presentation, identification of MDW, as a biomarker for disease severity and outcome, without additional testing is clinically relevant.

P25
Is Viscoelastometric Testing A Good Tool To Assess Hemostasis Of Covid-19 Patients?
Marion Bareille1, Michael Hardy2, Thomas Lecompte3, François Mullier1
1Université catholique de Louvain, Hematology laboratory, Namur Thrombosis and Hemostasis Center (NTHC), CHU UCL Namur, Yvoir, Belgium, 2Université catholique de Louvain, Anesthesiology department, Namur Thrombosis and Hemostasis Center (NTHC), CHU UCL Namur, Yvoir, Belgium, 3Departments of Medicine, Division of Angiology and Haemostasis, Geneva University Hospitals, and Geneva Platelet Group (GpG), Faculty of Medicine, University of Geneva, Geneva, Switzerland

Introduction: Infection by SARS-CoV-2 is associated with a high risk of thrombosis. The laboratory documentation of hypercoagulability and impaired fibrinolysis remains a challenge. Our main aim was to assess the potential usefulness of viscoelastometric testing (VET) to predict thrombotic events in COVID-19 patients according to the literature. Our secondary aims were: (i) to analyze the impact of anticoagulation and the methods used to neutralize heparin, (ii) to see whether maximal clot mechanical strength brings more information than Clauss fibrinogen, and (iii) to point out results from studies with enhanced fibrinolysis modified tests.
Methods: We performed a systematic search in PubMed and Scopus databases, until December 31st, 2020 (Figure 1). VETs methods and parameters, and patients’ features and outcomes were extracted.
Results: VET was performed for 1063 patients (893 ICU and 170 non-ICU, n= 44 studies). There was a huge heterogeneity concerning study design, the VET device (ROTEM, TEG, Quantra and ClotPro) and reagents (with non-systematic use of heparin neutralization by heparinase and/or polybrene), timing of assay, and definition of hypercoagulable state. The common findings were an increased clot mechanical strength mainly due to an excessive fibrinogen component with impaired to absent fibrinolysis, more conspicuous in the presence of an added plasminogen activator. This profile was associated with an uncontrolled thrombin generation despite a standard thromboprophylaxis. However, only 4 studies out of 16 that addressed this point found an association of VETs with thrombotic events. Functional fibrinogen assessed by VET showed a variable correlation with Clauss fibrinogen. Abnormal VET pattern tended to normalization after an enhancement in thromboprophylaxis. Notably, only 4 studies out of 25 using ROTEM reported data where heparin is neutralized by heparinase (HEPTEM).
Conclusions: The heterogeneity among VET studies and small sample sizes do not permit to point out an association between the ill-defined hypercoagulable state and thrombotic events.

P26
Evaluation Of The Lysis Timer To Investigate &Lsquo;Global Fibrinolysis Capacity&Rsquo; In Plasma 
Marion Bareille1, Michael Hardy2, Bernard Chatelain1, Thomas Lecompte3, François Mullier1
1Université catholique de Louvain, Hematology laboratory, Namur Thrombosis and Hemostasis Center (NTHC), CHU UCL Namur, Yvoir, Belgium, 2Université catholique de Louvain, Anesthesiology department, Namur Thrombosis and Hemostasis Center (NTHC), CHU UCL Namur, Yvoir, Belgium, 3Departments of Medicine, Division of Angiology and Haemostasis, Geneva University Hospitals, and Geneva Platelet Group (GpG), Faculty of Medicine, University of Geneva, Geneva, Switzerland

Introduction: There is currently no universal and standardized test available for plasma fibrinolytic activity. The ‘global fibrinolysis capacity’ (GFC) assay performed with the Lysis Timer instrument (SD Innovation; reagents and internal quality controls from Hyphen Biomed) does not have the euglobulin clot lysis time (ECLT) related preanalytical constraints. Our primary aims were i) to investigate the influence of the temperature used to keep the samples after sampling (4°C versus ambient temperature); ii) and that of a freezing-thawing cycle; and iii) to establish a preliminary reference range for GFC. Our secondary aim was to compare results of GFC with ECLT.
Methods: Influence of sample storage temperature and a freezing-thawing cycle were studied with fresh plasma samples from 30 healthy adult volunteers. Reference range was established on fresh plasma samples from 21 healthy adult volunteers. Method comparison was performed with 77 samples (healthy volunteers n = 34, patients with personal history of bleedings n = 43).
Results: There was no difference between GFC assessed in plasma samples processed on melting ice or at room temperature (p = 0.17), whereas GFC assessed in frozen-thawed plasma samples was prolonged when compared to fresh samples (p = 0.002) (Figure 1). Preliminary reference interval ranged from 29 (90%CI = [27-32]) to 50 (90%CI = [46-53]) minutes. GFC and ECLT were weakly correlated (ρ = 0.229, p = 0.04), with a poor agreement (κ = 0.03, 95%CI = [-0.154-0.221]) (Figure 2).
Conclusions: Sample storage temperature had no influence on the result. Reference ranges need to be established for fresh and frozen samples. Shorter than 1-hour reference interval enables the use of GFC in emergency cases. Finally, GFC and ECLT showed a weak correlation and a poor agreement, likely explained by plasma absence of fibrinolysis inhibitors such as PAI-1 in the euglobulin fraction.

P27
Bilateral Occlusion Of The Central Retinal Vein Due To Excess Of Factor Viii Level : A Case Report
Hanaa Bencharef
Chu ibn rochd casablanca , Casablanca, ON, Morocco

Introduction: Retinal vein occlusion (RVO) is an obstruction of the retinalvenous system, frequently linked to cardiovascular risk factors in eldery,and to thrombophilia in younger patients. In this article we report a case of bilateral central retinal vein occlusion (CRVO) ) in a young patientrelated to excess factor VIII levels.
Methods: A 22-year-old woman presented with a complaint of sudden decreased visual acuity in her both eyes, with severe high blood pressure and significant proteinuria. The patient had no particular pathological history. The best-corrected visual acuity was reduced to light perception in both eyes. A complete fundus assessment including retinal angiography showed bilateral occlusion of the central retinal vein with areas of ischemia. On the renal side, the patient had renal failure requiring hemodialysis sessions. A complete evaluation of the thrombophilia was carried out revealing an excess of the plasma level of factor VIII at 232% (Normal : 70%-140%)  - only abnormality of the balance sheet -, thus suggesting the diagnosis of thrombophilia due to an increase of the factor VIII. Panretinal photocoagulation were performed in both eyes. Additionally, to treat thrombophilia, the patient was referred to the hematology department for specialized therapeutic management.
Results:
Conclusions: An elevated factor VIII level is an independent risk of venous thrombosis.
Moreover, it is likely to influence the pathogenesis of central venous occlusion of the retina.These patients should be carefully evaluated to diagnose thrombophilia, including factor VIII, and initiate appropriate management as soon as possible.

P28
Hypofibrinogenemia Revealing Localized Intravascular Coagulopathy (Lic) In A Child With Venous Malformations: A Case Report
Hanaa Bencharef
Chu ibn rochd casablanca , Casablanca, ON, Morocco

Introduction: Venous malformations (VMs) are  vascular malformations present atbirth. Most of the themare unifocal, and only 1% are multifocal. These lesions, especially the diffuse and multifocal ones, are complicated by local intravascular coagulopathy (LIC) ; a consumptive coagulopathy typified by elevated D-dimer and decreased fibrinogen levels. The aim of this article is to report a case of multifocalVMs complicated with localized intravascular coagulation, revealed by low fibrinogen level.
Methods: We present 14-year-old boy with multifocal venous malformations, who presented a lateral cervical mass, rapidly increasing in size. Clinical examination showed a right submandibular swelling, measuring about 7 cm in diameter, with inflammatory skin, associated with other several painless soft masses present at the right axillary, dorsal, lower and right inguinal region . The patient underwent an exploratory cervicotomy with excision of a vascular-like latero-cervial mass, complicated by a postoperative hematoma for which he was reoperated to ensure hemostasis. Haematological investigations have essentially demonstrated an elevation of the D-dimer, with a very low fibrinogen level. These biological abnormalities occurring in association with venous malformations has been termed localized intravascular coagulation (LIC). Cervical CT scan revealed a tumor-like right latero-cervical formation, measuring 67x65x96.8 mm, well defined, heterodense, site of calcifications, heterogeneously enhanced after injection of PDC. 
Results:  the correlation between LIC and venous malformations is known, and it evolves, as the child ages and his venous malformations get larger. LIC is usually asymptomatic. However, it can cause swelling and pain due to thrombosis or bleeding within the lesions. This dysfunction of coagulationis characterized by the elevation of D-dimers and fibrin degradation products, low levels of fibrinogen, and sometimes minor-to-moderate thrombocytopenia, causing haemorrhagic and/or thrombotic complications which can lead to disseminated intravascular coagulopathy (DIC) and severebleeding during surgical procedures
Conclusions: : Recognizing LIC in patients with venous malformations is essential for detecting clinical and biological abnormalities, optimizing therapeutic management and thus preventing hemorrhagic and thromboembolic risks which may be life-threatening.

P29
Prevalence Of Lupus Anticoagulant In Covid-19 Outpatients
Nikolas Constantino, Rosana Cosentino, Claudio Mendes, Cristina Ito, Denis Mello, Valdir Aranda, João Guerra
HIAE, Sâo Paulo, Brazil

Introduction: The pandemic viral respiratory syndrome caused by the novel coronavirus (COVID-19) has already infected, by mid-October 2020, more than 5,224,362 victims in Brazil, totaling approximately 153,675 deaths. The aim of this study was to assess the prevalence of a possible risk factor for thrombotic complications in outpatients with COVID-19, the lupus anticoagulant (LAC), and compare it with individuals without the disease. The prevalence of D-dimer levels, as a thrombosis biomarker, was also verified.  
Methods: 122 citrated peripheral blood samples from outpatients with and without COVID-19 were analyzed in a core laboratory inside a Tertiary hospital. All samples were submitted to coagulometric tests to determine the presence of lupus anticoagulant, with two different methodologies: Diluted Russel's Viper Venom Test (dRVVT) and Silica Clotting Time (SCT), in accordance with the guidelines of the International Society on Thrombosis and Haemostasis (ISTH). Moreover, the data about patient's D-dimer tests were collected retrospectively from the databases of the computerized system. Statistical analysis was performed using Mann-Whitney, chi-square test, paired and non-paired t-Student test and Fisher Exact Test.
Results: Sixty nine samples from COVID-19 patients were included (median age 50,4 [5; 97] years, sex distribution [M:F=28:25] ) and compared with fifty-three non-COVID-19 individuals (median age 56,1 [12; 90] years, sex distribution [M:F=41:28]). Twenty-five (36,2%) samples from COVID-19 group tested positive for LAC, while positivity were observed in eleven (20,8%) samples from non-COVID-19 group (p=0,063). This study demonstrated a different behavior between the laboratory methods used, showing that the dRVVT hada highest positive rate in patients with COVID-19 (p=0,003).The number of patients with increased D-dimer was higher in the group without the disease (39,6% vs 36,2%, p=0,495).
Conclusions: We found a higher prevalence of LAC in patients with COVID-19, but this association was not statistically significant when compared to individuals from control group selected.  However, when the results are compared with the prevalence in general population, demonstrate a highest positive rate, mainly when is used dRVVT, possibly due to acute inflammatory phase interference in the complementary methods and the heterogeneity of aPL antibodies in these samples. Interestingly, the increased D-dimer levels was higher in the non-COVID-19 group, but without statistical significance.

P30
Influence Of The Covid-19 Pandemic On The Activity Of The Hematology Laboratory
Jose Ramon Furundarena, Carmen González, Alasne Uranga, Lurdes Agirrezabala, María Araiz, Arantza Agirre, Maialen Lasa, Nagore Argoitia, Laida Ondarra, Nerea Uresandi, Montse Lozano, Nerea Rodríguez, Adriana Sainz, Ana Arroyo
Donostia University Hospital, Donostia, Spain

Introduction: The pandemic caused by the SARS-CoV-2 virus has generated profound changes in health care. We have evaluated the influence of this pandemic on the demand for analytical tests.
Methods: The number of requests has progressively increased in recent years. In this study the demand for hematology laboratory tests has been collected over the last 2 years and the percentage of the change of the last year compared to the previous year has been calculated.
Results: Table 1 shows the annual request data for the different analytical tests of the hematology laboratory. During the year 2020 (year of the pandemic) there has been a widespread decline in demand for hematology analytical tests with the exception of demand for urgent hemostasis tests that have grown by 5-8 % and D-dimer requests that have increased by around 300 %. COVID-19 is associated with a higher risk for thrombotic complications and with an increase in D-dimer levels, which explains this increase in these analytical tests.        Table 1
  2019 2020 Difference (%)
Routine Hemograms 560099 465953 -16,8
Urgent Hemograms 85028 81893 -3,7
Total Hemograms 645127 547846 -15,1
Routine ESR 51068 40634 -20,4
Urgent ESR 1379 1065 -22,8
Routine PB morphology 16405 14027 -14,5
Urgent PB morphology 3634 3136 -13,7
Blood groups 10584 9467 -10,6
Direct Coombs 4886 3981 -18,5
Indirect Coombs 10554 10149 -3,8
Routine prothrombin time 129143 107349 -16,9
Urgent prothrombin time 56822 59791 5,2
Routine APTT 110098 91319 -17,1
Urgent APTT 52152 56438 8,2
Routine fibrinogen 6219 5219 -16,1
Urgent fibrinogen 3508 3800 8,3
Routine D-dimer 2435 10459 329,5
Urgent D-dimer 2673 9902 270,4
Thrombin time 981 755 -23,0
Clotting factors 1010 965 -4,5
Antithrombin 1076 789 -26,7
C Protein 1014 839 -17.3
S Protein 991 823 -17,0
Activated protein C resistance 811 688 -15,2
Lupus anticoagulant 2466 2103 -14,7
Collagen/Epinephrine closure time 181 113 -37,6
Collagen/ADP closure time 181 113 -37,6

Conclusions: The COVID-19 pandemic has conditioned a clear decrease in the demand for hematology laboratory tests except for the demand for D-dimer and urgent hemostasis tests that have experienced a clear increase in relation to thrombotic risk associated with the virus.

P31
Infectious Mononucleosis Manifesting With Cold-Type Autoimmune Hemolytic Anemia
Kevin Ginnebaugh, Harshita Mehrotra, Sunny R K Singh, Zaher Otrock
Henry Ford Hospital, Detroit, MI, United States

Introduction: Ebstein-Barr virus (EBV), the causative agent of infectious mononucleosis (IMN), is a widely disseminated human pathogen with more than 90% of adults being seropositive. Patients with IMN can develop cold-type autoantibodies; however, cold-type autoimmune hemolytic anemia (AIHA) is a rare manifestation of infectious mononucleosis. Here, we describe a case of cold-type AIHA with transaminitis induced by IMN.
Methods: Case report
Results: 20-year-old female presented to the emergency room with fever, sore throat, nausea and vomiting. She was febrile (temp 38.2 deg C) and tachycardic (pulse 115/sec); her workup was consistent with IMN with a positive monospot test and mild anemia (Hgb 11 g/dL). She was discharged on supportive treatment. However, patient returned after 3 days with left sided abdominal pain and icteric sclera. Her physical examination revealed splenomegaly, and laboratory workup was significant for hemolytic anemia, transaminitis, and hyperbilirubinemia (Table). EBV viral capsid antigen IgM was positive. Ferritin level was significantly high. Direct antiglobulin test (DAT) was positive for C3d only, and cold hemagglutinins were positive with a high thermal amplitude of 30 deg C. Peripheral blood smear was noticeable for leukocytosis with atypical lymphocytes and marked anemia with aggregates of red blood cells (RBC). Further workup for autoimmune diseases, malignancy, and other infections was negative. Patient was managed supportively during hospital admission, and she required one unit of packed RBCs transfused with warming. She was discharged 1 weeks later, and at 1 month follow up she showed complete clinical improvement and resolution of abnormal laboratory workup.
Conclusions: Our patient had AIHA in the form of secondary cold agglutinin disease, which is a very rare complication of EBV. She presented with significant anemia, hemolysis and transaminitis. Cold agglutinins with high thermal amplitude were identified. She was managed supportively, and recovered very well. Ferritin, which is an acute phase reactant, was a good marker of disease and recovery.

P32
Determination Of The Reference Intervals Of The Willebrand Factor At Center For Hemobiology-Blood Transfusion
kahina Guenounou-Ghemmour1, Soraya Hadjali -Saichi 2, Issam Frigaa3
1Center for Hemobiology-Blood Transfusion, Algiers, Algeria, ALGERIE, Algeria, 2Center for Hemobiology-Blood Transfusion, Algiers, Algeria, ALGERIE, Algeria, 3Center for Hemobiology-Blood Transfusion, Algiers, Algeria, ALGERIE, Algeria

Introduction: The factor Willebrand is a glycoprotein representing an important Maillon in primary hemostasis, its deficiency is responsible for Willebrand disease, its diagnosis is by thedetermination of plasma levels of VWF. Variations in the intervals between laboratories are therefore frequent, and reference intervals are to be determined for each laboratory. Objectives: Determination of reference intervals for the cofactor activity of Ristocetin (STA® - VWF:RCO on STARMAX2), VWF: Ag Liatest and ratio VWF: Rco / VWF: Ag

     
Methods: Our study involved 146 controls: 29 children ,44 adult women and 73 adult men. 51 were from Group O, 50 were from Group A, 26 were from Group B and 19 were from Group AB.
Results: The study of the reference intervals of VWF : Rco groups O and not O shows a significant difference between these two groups (p 0.0001). The VWF : Rco reference interval of group O is [42-114] %. The range of non O is [69-160] %. Blood groups A, B and AB show no significant differences between them (p > 0.05).
The study of the antigen Willebrand VWF: Ag Liatest also showed a significant difference between the groups O and not O (p < 0.05). The lower limit of the ratio VWF: Rco/ VWF: Ag was 0.76 in children, 0.79 in adult women and 0.68 in adult men respectively with a non-significant difference (p > 0.05) in all three groups.
Conclusions: This work allowed us to determine the VWF reference intervals of the parameters: VWF: Rco, VWF: Ag, and the ratio VWF: Rco/VWF: Ag Liatest.  

P33
Comparison Of Two Techniques For The Determination Of Willebrand Factor Activity
kahina Guenounou-Ghemmour, Soraya Hadjali -Saichi , Issam Frigaa
Center for Hemobiology-Blood Transfusion, Algiers, Algeria,, ALGERIE, Algeria

Introduction:  Willebrand factor is a glycoprotein synthesized by endothelial cells and megakaryocytes.
OBJECTIVE: Comparison between two techniques for the determination of the cofactor activity of ristocetin , an automated technique on STARMAX2 (STA® - VWF:RCO) and standard aggregometric method using the HYPHEN- BIOMED platelets.
Methods: Our study involved 61 patients. The results of the aggregate test were rendered by exploiting the velocity data in manual and automatic mode on the thrombo -TA-8V aggregator.
Results: The study of the correlation shows a positive and highly significant correlation r=0.771 (p0.0001) between the VWF : RCO on STARMAX2 and VWF : RCO by exploiting the data of manual velocities , The concordance analysis by the Bland-Altman diagram, shows that there is a positive bias at 8.16%, not significant .The specificity of the VWF: RCO method on STARMAX2  was better at 96.07% followedby the VWF:RCO method by exploiting the data
of manual velocities( specificity at
84.43%).the VWF:RCO method using the automatic velocity data showed poor specificity at 41.17%.The sensitivity of the three techniques was 100%. Our work, as well as several studies, haveshown that the coefficients of variation in the aggregate technique were higher than theautomated technique which could thus be the source of the systematic transition from anormal to a pathological zone forthe same patient depending on the technique used
Conclusions: In Algeria the standard aggregometric method remains the most used technique. For laboratories that do not have the automated technique, it could be recommended to render the results of the Willebrand activity using the manually calculated velocities.    

P34
Comparison Of Three Anti -Xa Activity Measuring Reagents
Soraya Hadjali-Saichi, Kahina Guenounou-Guemmour, Issam Frigaa
Center for Hemobiology-Blood Transfusion Faculty of Medicine of Algiers , Algiers, Algeria

Introduction: The anti-Xa activity specifically determines the anticoagulant activity of the unfractionated  heparins  (HNF) by measuring the heparin-related antithrombin’s ability to inhibit a single enzyme, factor Xa (FXa). Recent improvements in the automation, cost-effectiveness and accessibility of the test to clinicians have made the Anti-Xa test a part of daily clinical practice in many institutions.The objective of this study is the comparison of the results of the anti-Xa activity measurement by three different marketed reagents available in our laboratory: STACLOT®HEPARIN (chronometric test), STACHROM ®HEPARIN and STA®-LIQUID ANTI-Xa (chromogenic tests) ; to evaluate the variability of measurement between them.
Methods: Samples from 60 patients  under HNF were collected. Their citrated plasmas, obtained after centrifugation, were analysed within two hours of sampling, using three techniques at the same time. For comparison of results, the statistical tests applied are: the Lin concordance coefficient (CClin) and the Bland-Altman difference diagram.
Results: A match between the results obtained by the chronometric and chromogenic assay, was found( CCLin  for  STACLOT® HEPARIN versus STA-®Liquid Anti-Xa = 0,84  and CCLin  for STACLOT® HEPARIN versus  STACHROM ®HEPARIN = 0,71), with an over-estimation of the results by the STACLOT® HEPARIN test( d = 0,06 UI/ml for  STACLOT® HEPARIN versus STA-®Liquid Anti-Xa and d = -0,02UI/ml for STACLOT® HEPARIN versus  STACHROM ®HEPARIN). These differences between the results are clinically non-significant ( 0.2 IU/ml).
Conclusions:  The results obtained by the various statistical tests applied show a good match between STACLOT® HEPARIN, STACHROM ® HEPARIN and STA®-Liquid Anti-Xa.  This study could be considered as a good starting point towards the standardisation of the anti-Xa tests of the different laboratories.

P35
Acquired Glanzmann'S Thrombasthenia Associated With Chronic Lymphoblastic Leukemia (Case Report)
jing jin1, James Zehnder2
1Department of Special Coagulation, Clinical laboratories, Stanford Hospital & Clinics, Palo Alto, CA, United States, 2Department of Pathology, School of Medicine, Stanford University, Stanford, CA, United States

Introduction: Acquired Glanzmann's thrombasthenia (GT) is an uncommon event in association with leukemia usually due to autoantibodies against normally expressedαIIbβ3. The authors describe a patient with chronic lymphoblastic leukemia (CLL) who had developed easy bruising, hematomas and post-surgery severe bleeding, in association with abnormal progressively worsening platelet function assays in the setting of progressive lymphocytosis over 6 years while untreated. An acquired GT due to CLL was diagnosed. After treatment with obinutuzumab, his CLL went into sustained complete remission with normalization of  platelet function assays and resolution of bleeding. 
Methods: Platelet function was assessed by platelet function screen (PFA-100) and light transmission platelet aggregation (LTA) at multiple timepoints before and following therapy of CLL as indicated in the Figure.
Results: Patient’s peripheral blood flow cytometry showed an abnormal CD19 positive B-cell population coexpressing CD5 and CD23 in 2013. The patient experienced progressive lymphocytosis with white blood cell count (WBC) increasing from 14.0 K/uL in 2013 to 140.0 K/uL in 2019 over a period of 6 years. The patient’s course was  complicated by easy bruising and postoperative bleeding following  hernia repair. Two  platelet function screens (PFA-100) and two platelet aggregation studies by LTA had been performed. The first PFA-100 in 2018 had a prolonged closure time to  Collagen/Epinephrine (EPI) only, while the second PFA -100 in 2019 showed prolonged closure with both Collagen/EPI and Collagen/ADP  cartridges. The initial LTA study in 2013 showed reduced response to EPI only while the study in 2018 revealed a global decreased response to ADP, EPI, arachidonic acid, collagen and TXA2 analogue U46619, with normal aggregation to ristocetin, consistent with acquired Glanzmann’s thrombocythemia.  In September of 2019, obinutuzumab (anti-CD20 therapy) was started for his CLL with compete response.  His WBC count and PFA-100 tested one month after the treatment were all normal. His bleeding symptoms resolved and a follow- platelet aggregation study had normalized.
Conclusions: We evaluated a rare case of acquired Glanzmann's thrombasthenia associated with CLL. Laboratory platelet function assays closely correlated with patient’s clinical manifestation and CLL treatment. Although rare, awareness of this hemorrhagic diathesis as a possible sequelae of lymphoproliferative disorder should encourage clinical vigilance of these patients.

P36
Evaluation Of Aggreguide A-100 With Platelet Function Tests And Thromboelastrography In Patient With Ischemic Stroke.
Young Woo Kim1, Hey Kyung Lee2, Hyunjung Kim2
1Department of Neurosurgery, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea , Seoul, South Korea, 2Laboratory Medicine, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea , Seoul, South Korea

Introduction: This study investigated the clinical utility and functional comparison of four platelet reactivity measured by four platelet function tests in patients with ischemic stroke and taking clopidogrel.
Methods: From July 2018 to May 2019, thirty-five patients who developed acute ischemic stroke were enrolled. Patients received an initial loading dose of 300 mg clopidogrel followed by a 75 mg daily maintenance dose. Platelet aggregation was measured by newly introduced AggreGuide A-100, VerifyNow P2Y12 (VerifyNow), platelet function analyzer P2Y (PFA) and light transmission aggregometry (LTA). Five parameters were obtained using Thrombelastography (TEG) (Haemonetics, USA). Stroke type, region and other clinical data were obtained from chart review. Patients were monitored for one year for recurrence, bleeding complication and death.
Results: Artherosclerotic type was 25 patients, and small artery disease type was 10 patients. The stroke recurrence, bleeding complication and death did not occurred during the monitoring period. High on-treatment platelet reactivity (HTPR) rate was 18.2% in VerifyNow, Aggreguide 27.3%, PFA 7.2%, LTA 9.1%. Concordance rates were as follow; VerifyNow and PFA 92.6%, VerifyNow and Aggreguide 60.6%, VerifyNow and LTA 72.7%, PFA and Aggreguide 70.4%, PFA and LTA 88.9% and, Aggreguide and LTA 91.7%. The correlation between methods was not good (The correlation coefficient r was from 0.176 to 0.673). The clinical and other coagulation markers was described in Table 1. Blood viscosity and NIH stroke scale (NIHSS) were associated with HTPR on Aggreguide, and TEG K-time,blood viscosity and NIHSS were associated with HTPR on VerifyNow.
Conclusions: In this study, stroke recurrence was absent, therfore HTPRs were compared. The incidence of HTPR was lowest in PFA and highest in Aggreguide. Agreement between tests was good between VerifyNow and PFA, Aggreguide and LTA, and PFA and LTA. However correlation was not good, the data are not interchangeable.

Best Student Presentation Award Finalist
P37
The Aptt Clot Waveform Analysis Can Distinguish Between Hemophilia- A Non-Inhibitors And Hemophilia- A With Inhibitor: A Pilot Study From India 
Khalid Abdul Mannan1, Narender Kumar 2, Jasmina Ahluwalia3, Chander Hans4, Gaurav Prakash5, Richa Jain6
1Dept. of Hematology, PGIMER, Chandigarh, India, 2Dept. of Pathology, PGIMER, Chandigarh, India, 3Dept. of Hematology, PGIMER, Chandigarh, India, 4Dept. of Hematology, PGIMER, Chandigarh, India, 5Dept. of Internal Medicine, PGIMER, Chandigarh, India, 6Dept. of Paediatrics, PGIMER, Chandigarh, India

Introduction: Hemophilia A patients have isolated activated partial thromboplastin time (aPTT) prolongation.  The presence or absence of inhibitors cannot be suggested only on the basis of aPTT and requires the inhibitor screening based mixing study and thereafter Bethesda Inhibitor assay.  Clot Waveform Analysis (CWA) is the extended study of the slope generated by an optical detection system during aPTT or PT. Multiple quantitative and qualitative parameters of aPTT CWA provides valuable information in addition to clotting time.  The present study was planned to assess whether the CWA can differentiate  between Hemophilia A  samples with and without inhibitors.
Methods: A total number of 66 samples from 55 patients with prolongation of aPTT were recruited. The CWA parameters (Figure 1) were noted manually   on the ACL-TOP 500 Coagulation analyser screen using mouse pointer. In all these samples PT, aPTT, fibrinogen, mixing study, Inhibitor screen, Bethesda assay, FVIII assay, FIX assay, FXI assay, vWF: Ag and vWF:GP1bR assay were performed.
Results: There were 50 samples  of HA non-inhibitor and 16 samples of HA with inhibitor.   Each disease was divided arbitrarily on the basis of aPTT in to category 1 (< 70 sec), category 2 (71-100 sec), and category 3 (>100 sec), as these cases cannot be differentiated on the basis of factor level due to the presence of inhibitors. In each category the CWA parameters were compared using Mann-Whitney U-test.   In the category 1, the quantitative parameters with a statistically significant difference are: Acceleration peak time (1A) (p=0.030), Velocity peak time (1B) (p=0.013), Time for ½ height fibrin formation (1C) (p=0.019), Height of acceleration/Secondary derivative [+] peak (2A) (p=0.030).  The category 2 had the following statistically significant quantitative parameters: 1C (p=0.037), 2A (p=0.048), and height of acceleration/Secondary derivative [-] peak (2B) (p=0.048).
Conclusions: The aPTT-CWA can distinguish between Hemophilia A cases (with & without inhibitor) The CWA parameters that can help in differentiation are time for ½ height fibrin formation in the fibrin curve; and in the 2º DC, the height of acceleration [+] peak if the aPTT is less than 100 sec.  

P38
Clinicopathological Parameters In Haemophilia A Patients With And Without Inhibitors
Debadrita Ray, Narender Kumar, Jasmina Ahluwalia, Reena Das, Chander Hans, Arihant Jain, Pankaj Malhotra, Deepak Bansal
PGIMER, Chandigarh, India

Introduction: Factor VIII (FVIII) replacement is the mainstay of treatment in haemophilia A (HA) patients. However, FVIII replacement may lead to the development of inhibitors. While a vexing clinical problem, some observations suggest that the presence of inhibitors may not necessarily portend a higher bleeding risk. We looked at the clinicopathological profiles and bleeding risk in HA patients with and without inhibitors.    
Methods: Clinical details and laboratory findings of consecutive HA patients attending a north-Indian tertiary-care centre from 2010-2020 were retrospectively reviewed.
Results: Out of 600 HA patients, inhibitors were detected in 35 patients (5.8%). Median age of patients with inhibitors was 19 (4.7-31.2) years and family history of HA was present in 20 patients (57.1%). Most patients with inhibitors had history of transfusion with FVIII alone (54.3%) or a combination of FVIII concentrate and other blood products (42.8%). Only 1 patient (2.85%) with inhibitor had no history of FVIII concentrate transfusion. Patients with blood-group A had a higher risk of developing inhibitors (OR-3.02; 95% CI:1.06-8.82, p-0.04). Overall hemarthroses was the commonest site of bleeding seen in 91.4% and 87.7% patients with and without inhibitors, respectively. Intracranial bleed was significantly more frequent in patients with inhibitors compared to those without inhibitors (20% vs 4.1%; p-0.03). Time dependent (TD) and immediately acting (IA) inhibitors were seen in 60% and 40% patients, respectively. High-titre (>5 BU/mL) and low-titre inhibitors (< 5 BU/mL) were detected in 28 (80%) and 7 (20%) patients, respectively. Median inhibitor titre in patients with high and low inhibitor titres were 35.09 and 4.4 BU/mL, respectively. There was no significant difference in age among patients with high and low titre inhibitors (21 vs 15.5 years; p=0.29). No difference in type of inhibitor was seen between those with high [TD=15 (53.6%), IA=13(46.4%)] or low titre [TD=6 (85.7%), IA=13(46.4%), p=0.2] inhibitors.
Conclusions: Prevalence of inhibitors in our cohort was 5.8% and most had high-titre, time dependent inhibitors. These patients may have a higher risk of intracranial bleeding.

P39
Discrepancy Between Genotype And Phenotype For Factor V Leiden Mutation In Recipients Of Liver And Stem Cell Transplantation
Joseph Rigano
Alfred Health, Melbourne, Australia

Introduction: Activated protein C resistance (APCR) was first described in 1993 (Dahlbäck et el.) and is the most common hereditary risk factor for venous thromboembolism (VTE) in the Caucasian population. The majority of cases result from factor V Leiden (FVL) mutation (R506Q) which causes activated factor V to be resistant to APC inactivation resulting in a hypercoagulable state. Thrombophilia testing for APCR of liver synthesised FV and FVL mutation in peripheral blood leukocyte DNA can lead to discrepancies between genotype and phenotype in transplantation recipients. This report describes 2 cases; acquired APCR without FVL mutation following liver transplantation (LTX) and inherited APCR without FVL mutation following stem cell transplantation (SCTX) both associated with deep vein thrombosis (DVT).
Methods: The first case was a 68-year-old male who presented with DVT nine months post LTX for liver cirrhosis caused by HCV. The second case was a 42-year-old female who presented with CRT four months post SCTX for AML. Both patients reported no history of thrombosis prior to transplantation. Thrombophilia assays (protein C, protein S, antithrombin, APCR, lupus anti-coagulant, anti-cardiolipin and anti-β2-glycoprotein I antibodies) were performed using HemosIL® reagents on the ACL TOP CTS 500 and AcuStar analysers (Instrumentation Laboratory; Werfen). Molecular thrombophilia assays for FVL and prothrombin gene (G20210A) mutations were performed using Qiagen Rotor-Gene by PCR and HRM analysis.
Results: Protein C, protein S and antithrombin levels were normal and assays for anti-phospholipid antibodies were negative. FVL and prothrombin gene mutations were not detected. APCR was detected in both patients with ratios of 1.75 and 1.64 for case one and two respectively (normal APCR ratio 2.2–3.3). In case one, APCR was detected in donor liver and recipient’s peripheral leucocyte DNA lacked FVL mutation. In case two, APCR was detected in recipient’s liver and peripheral leucocyte DNA of donor lacked FVL mutation. Since both patients reported no history of thrombosis, thrombophilia testing was never indicated. It has been established that LTX and SCTX recipients are at risk of VTE. Therefore, consideration should be given to thrombophilia testing of LTX and SCTX donors and recipients which may indicate anticoagulation to prevent additional morbidity.
Conclusions: Here reported were two cases of genotype and phenotype discrepancies for FVL mutation in recipients of LTX and SCTX both associated with DVT.

P40
Investigation Of Methods Thrombin&Rsquo;S Activation In Technolgy Of Purification 
Nataliia Shurko, Taras Danysh , Vasyl Novak
State Institution «Institute of Blood Pathology and Transfusion Medicine NAMS of Ukraine», Lviv, Ukrenia

Introduction: The purified Thrombin’s concentrate is used in medical and clinical practice as a therapeutic (for local cessation of bleeding from small capillaries and parenchymal organs) and diagnostic (for coagulation studies) agent.  The process of purification of Thrombin usually requires a plurality combination of steps purification and activation. Aim:to select the optimal conditions for the activation of Prothrombin to Thrombin in the process of purification. 
Methods: The Cohn fraction II+III was used as initial material. The PPSB, the plasma derivative which contains the vitamin K-dependent clotting factors, were obtained by combining methods of adsorption/precipitation, suspension and centrifugation. The Thrombin actvity was assessed by thrombin time.
Results: The effect of different reagents on the activity of Thrombin was investigated. In the process of activation of Thrombin we added the following reagents to the solution of PPSB: 40 mM CaCl2; 15 % PEG 6000 and 5 mM CaCl2; 15 % PEG 4000 and 5 mM CaCl2; 15 % PEG 1500 and 5 mM CaCl2; 15 % PEG 400 and 5 mM CaCl2; solution of thromboplastyn.The determination of Thrombin activity was performed immediately, after 1hour, 12, 48 and 120 hours. The highest level of Thrombin activity was obtained on the 5th day of activation using 15% PEG 4000 and 5 mM CaCl2(1005.48±30.23 NIH/ml). When using 40 mM CaCl2 or solution of 15% PEG 6000 and 5 mM CaCl2, the level of Thrombin activity was about 840 (862.48±16.43 NIH/ml and 820.08±9.33 NIH/ml, respectively).The similar data wereobtained usingsolution of  thromboplastyn or 15% PEG 1500 and 5 mM CaCl2 as activators. The level of Thrombin activity was about 415 (420.19±14.03 NIH/ml  and 410.88±17.35 NIH/ml, respectively). It should be noted that the activation process of 15% PEG 400 and 5 mM CaCl2 was the lowest (252.48±9.76 NIH/ml).
Conclusions: Was found that 15% PEG 4000 and 5 mM CaCl2 to provide the highest degree of thrombin activation.The activated Thrombin solution was used for further purification by the method dye-ligand affinity chromatography.

P41
A National Reference Laboratory&Rsquo;S Review Of Test Ordering Patterns And Positivity Rates During The Covid-19 Pandemic
DA Stephens1, RA Crist1, GM Rodgers1,2, KJ Smock1,2, KA Moser1,2
1ARUP Laboratories, SLC, UT, United States, 2University of Utah Department of Pathology and Medicine, SLC, UT, United States

Introduction: The COVID-19 pandemic had a significant impact on the field of laboratory medicine. Many procedures were delayed to reduce the burden on the medical system. This led to a nationwide change in the testing associated with those procedures. In addition, patients with COVID-19 had increased thrombotic risk and many received anticoagulation. In this study, we aimed to summarize changes to ordering and result patterns in lupus anticoagulant and heparin-induced thrombocytopenia testing that were observed in our national reference laboratory. 
Methods: We examined 2019 and 2020 test volumes and positivity rates for two assays, the Lupus Anticoagulant Reflexive Panel (LA) and the Serotonin Release Assay (SRA) for heparin-induced thrombocytopenia (HIT) due to their connection to COVID-19 complications or therapies.
Results:   LA volumes dropped during March and April 2020 (-17% and -39%, respectively), then recovered to normal levels by June 2020 without significant differences in average monthly volume for the remainder of 2020 versus 2019. The LA positivity rate showed a -25% change in June 2020 (7%) versus 2019 rates (9%) and remained lower for the rest of the year.  On average over the entire year, there was a -20% difference in LA positivity rate in 2020 versus 2019.     2020 SRA volumes were similar to 2019 volumes during the first half of the year, started to increase in August 2020 (9% difference), and remained high. SRA positivity rates dropped in March and April 2020 (-16% and -50% difference, respectively), and returned to normal levels in September 2020.
Conclusions: Our high volume reference laboratory observed significant changes in test ordering patterns and positivity rates for certain tests during the pandemic.   Although speculative, changes in LA positivity rate may be due to differences in patient populations being tested (increased testing in COVID-19 patients and decreased testing in other populations). Although COVID-19 patients have increased thrombotic risk, it is not thought to be due to anti-phospholipid antibodies. The changes in SRA positivity rate may have been due to the cancellation of surgeries with high HIT risk and relative increase of medically ill patients receiving prophylactic heparin. In both cases, trends are likely due to an increased awareness of the thrombotic issues seen in COVID-19 patients.

P42
Aprotinin Is Not An Idle Bystander In The Interpretation Of Rotational Thromboelastometry Analysis.
Frederick Verbeke, Pieter De Kesel, Katrien Devreese
Ghent University Hospital, Ghent, Belgium

Introduction: Aprotinin has an antifibrinolytic effect and is indicated for prevention of bleeding in high-risk adults receiving coronary bypass surgery. Previously, the prolonging effect of aprotinin on the activated partial thromboplastin time, activated clotting time, and clotting time (CT) measured by rotational thromboelastography was demonstrated. Here, the effect of aprotinin on rotational thromboelastometry (ROTEM) INTEM and HEPTEM CT was for the first time investigated in the presence of increasing unfractionated heparin concentrations.
Methods: Citrated whole blood from 5 healthy volunteers was spiked with 250 kallikrein inhibitor units (KIU)/mL aprotinin (Trasylol®, 500 000 KIU/50 mL, Nordic Pharma, Antwerp, Belgium) in the presence of heparin (Heparin Leo®, 5 000 international units (IU)/mL, Leo Pharma, Ballerup, Denmark) corresponding to anti-Xa levels up to 1.25 IU/mL anti-Xa. The CT INTEM and HEPTEM were determined on a ROTEM Delta (Werfen, Barcelona, Spain) and statistical analysis was performed by multiple regression or an one-way ANOVA with post hoc Tukey’s honestly significant difference test (SPSS 27, IBM, Armonk, NY, USA).
Results: Aprotinin significantly prolonged both the CT (i.e. dependent variable) INTEM in a heparin concentration dependent way (R² = 0.623, F(2, 67) = 58.119, p < 0.0001) and CT HEPTEM  independent of the heparin concentration (R² = 0.245, F(2, 67) = 12.207, p < 0.0001) with heparin and aprotinin concentrations as independent variables. In the absence of heparin, aprotinin prolonged the CT INTEM by 33% (95% confidence interval (CI): 21%; 45%), and the CT HEPTEM by 28% (95% CI: 20%; 37%). At different concentrations of heparin, CT INTEM and CT HEPTEM were on average 55% (range: 41%; 72%) and 51% (range: 41%; 69%) prolonged, respectively. The ratio CT HEPTEM/CT INTEM, a measure used to evaluate the (residual) effect of heparin, was decreased by aprotinin (p < 0.001) by on average -16% (range: -21%;-10%) in the presence of heparin.
Conclusions: Aprotinin prolonged the CT INTEM and CT HEPTEM, and decreased the ratio CT HEPTEM/CT INTEM, resulting in an overestimation of the heparin effect as evaluated via the ratio CT HEPTEM/CT INTEM.

P43
Point-Of-Care Kaolin Activated Clotting Time In Heparin Monitoring: Mind The Aprotinin!
Frederick Verbeke, Pieter De Kesel, Katrien Devreese
Ghent University Hospital, Ghent, Belgium

Introduction: Point-of-care heparin monitoring via activated clotting time (ACT) is frequently performed during coronary bypass surgery to adjust heparin dosing. To reduce blood loss during coronary bypass surgery, aprotinin can be additionally administered. In the early 1990’s, the effect of aprotinin on ACT was described by several authors, and especially celite-activated ACT resulted in more prolonged ACT compared to kaolin-activated ACT. According to the manufacturer however, the i-STAT Kaolin ACT cartridge can be used to monitor heparin therapy in the presence of aprotinin. The effect of aprotinine on the i-STAT kaolin ACT in the presence of increasing unfractionated heparin concentrations (UFH) was therefore examined.
Methods: Citrated whole blood from 5 healthy volunteers was spiked with 250 kallikrein inhibitor units (KIU)/mL aprotinin (Trasylol®, 500 000 KIU/50 mL, Nordic Pharma, Antwerp, Belgium) in the presence of heparin (Heparin Leo®, 5 000 international units (IU)/mL, Leo Pharma, Ballerup, Denmark) corresponding to anti-Xa levels up to 1.25 IU/mL anti-Xa. Immediately before ACT analysis, the citrated whole blood was recalcified by adding 3.2 µL 2.5 M CaCl2 to 500 µL of whole citrated blood. ACT was measured using the i-STAT Kaolin ACT device and cartridges (Abbott, Chicago, IL, USA) and statistical analysis was performed by multiple regression or an one-way ANOVA with post hoc Tukey’s honestly significant difference test (SPSS 27, IBM, Armonk, NY, USA).
Results: ACT was defined as the dependent variable and aprotinin and heparin concentration were defined as independent variables (R² = 0.623; F(2, 67) = 68.925, p < 0.0001). Both aprotinin (p = 0.001) and heparin (p < 0.0001) significantly prolonged the ACT. In the absence of heparin, aprotinin prolonged the ACT by on average 7.4% (95% confidence interval: -2.2%; 17.0%; p = 0.480). When heparin at different concentrations (up to 1.25 IU/mL anti-Xa) was present, the ACT was on average prolonged by 7.9% (range: 4.9%; 10.2%) due to aprotinin.
Conclusions: A kaolin activated ACT was on average prolonged in vitro by aprotinin by 7.9% in the presence of heparin. Despite  kaolin ACT is less affected by aprotinin during heparin monitoring, our results indicate that caution is still mandatory when performing kaolin ACT in the presence of aprotinin to not overestimate the heparin effect. 

P44
The Haemoglobin Level As An Additional Criterion For The Severity Of Chronic Thromboembolic Pulmonary Hypertension
Ekaterina Zolotova1, Yulia Zhilenkova1, Maria Simakova2, Tatyana Vavilova 1, Olga Moiseeva2
1Almazov National Medical Research Centre, Faculty Department of Laboratory Medicine and Genetics, Saint-Petersburg, Russia, 2Almazov National Medical Research Centre, Research Department of the Non-Coronary Myocardial Diseases, Saint-Petersburg, Russia

Introduction: It was shown that low (< 140 g/L) and high (>160 g/L) levels of haemoglobin in cardiovascular patients are associated with worse outcomes. The patients with chronic thromboembolic pulmonary hypertension (CTEPH) are characterised with secondary compensating the hypoxemia erythrocytosis. The aim is to reveal the connection between haemoglobin levels and risk factors of CTEPH patients one-year mortality.
Methods: 120 CTEPH patients who had been enrolled in Almazov National Medical Research Centre CTEPH Registry between 2013 and 2020 were included in the study: 64 (53%) male and 56 (47%) female. The median age of patients was 56.7 ± 15.5. The study was complied with the Declaration of Helsinki. CTEPH diagnosis was verified according to European Society of Cardiology guidelines 2009, 2015. Anaemia was defined by WHO criteria (haemoglobin < 120 g/L in women and < 130 g/L in men). Characteristics of complete blood count were collected. 
Results: The mean haemoglobin concentration was 149.5 g/L [135.3; 159.9], and mean red blood cells amount was 5.18 [4.78; 5.58]. The prevalence of anaemia in CTEPH patients was 16% (19 patients). When dividing patients by the severity of CTEPH in the group with functional class (FC) I-II, anaemia was less common - 16.7% of patients, in the case of a more severe course (FC III-IV) – 22.8% (p=0.146). The weak positive correlation between haemoglobin level and such risk factor of adverse CTEPH outcome during the year as mixed venous oxygen saturation (SvO2) (59.5 ±10%, 0.331, p< 0.01) Fig.1 and distance in 6-Minute Walk Test (294 ± 122 m, 0.242, p< 0.05) was obtained.
Conclusions: The significance of haemoglobin level for clinical status of CTEPH patient was shown. Further studies are required to clarify the mechanism of anaemia in these patients and assessment of haemoglobin levels as an independent predictor of an adverse CTEPH outcomes.

P45
Thymoquinone Induces Jak/Stat Negative Regulators Re-Expression And Inhibits Cell Proliferation With Apoptosis Induction In Acute Myeloid Leukemia Cells
Futoon Al-Rawashde1, Hamid Al-Jamal2
1Universiti Sultan Zainal Abidin, KUALA TERENGGANU, Malaysia, 2Universiti Sultan Zainal Abidin, KUALA TERENGGANU, Malaysia

Introduction: Constitutive activation of JAK/STAT signaling pathway plays a crucial role in tumor cell survival and proliferation in acute myeloid leukemia (AML). SHP-1, SOCS-1,SOCS-3 tumor suppressor genes are the negative regulators of the JAK/STAT pathway. Thymoquinone (TQ) is the major biologically active compound of Nigella Sativa (black seeds) that has shown anticancer properties in several cancers. This study aimed to investigate the effect of TQ on MV4-11 AML cells. For this purpose cytotoxicity, apoptosis, cell cycle assays and gene expression of the JAK/STAT negative regulators were investigated in MV4-11 cells after treatment with TQ compared to the untreated cells.
Methods: Acute leukemic cell line (MV4-11) was treated with TQ. Cytotoxicity of TQ was determined andapoptosis rates and Cell Cycle distribution analysis were determined using Flow Cytometry. Quantitative reverse transcription polymerase chain reaction (qRT-PCR) was performed to evaluate the messenger RNA expression of SHP-1, SOCS-1,and SOCS-3.
Results: TQ caused MV4-11 growth inhibition in adose and time dependent manner. The IC50 values for TQ in MV4-11 cell line were 7.8 ±1.6, 5.5 ±1.3 and 3.8 ±0.96 μM, for 24, 48, and 72 hours, respectively. TQ induced apoptosis in MV4-11 cell line in a dose and time dependent manner. Apoptosisof MV4-11 cells was associated withdecreased G2/M peak, while these cells were arrested in G0/G1 phase in a dose and time dependent phenomenon.Furthermore, exposure of MV4-11 cells to TQ induced upregulation of SHP-1, SOCS-1, and SOCS-3 accompanied with downregulationof JAK2,STAT3,STAT5A andSTAT5B.
Conclusions: These findings revealed that TQ mediates its apoptotic effects and cell cycle arrest on AML by upregulation of the JAK/STAT signaling pathways' negative regulators, which needs further analysis of the anti-leukemic effect.   

Best Student Presentation Award Finalist
P46
Lymphocyte Subset Count And Outcome Of Covid-19 Patients Admitted To A Highly Complex Hospital In Southern Brazil.
Ana Azambuja1, Douglas Camara2, Yara Carolina Schluga3, Julie Lillian Pimentel Justus4
1Universidade Federal do Paraná, Curitiba, Brazil, 2Universidade Federal do Paraná, Curitiba, Brazil, 3HC-UFPR, Curitiba, Brazil, 4HC-UFPR, Curitiba, Brazil

Introduction: Introduction: Lymphopenia is one of the laboratory markers of SARS-CoV-2 infection considered a factor of poor prognosis in relation to the severity of COVID-19 disease. Studies indicated that evaluation of lymphocyte subsets, such as CD4 and CD8 T-cells can provide information for the severity of disease and convalescence. 
 
Methods: Methods: 121 patients attended at our hospital between April and June 2020 were analyzed. Patients with bacterial pneumonia and mild or moderate COVID-19 patients were not included. Routine multiparametric flow cytometry (MFC) from peripheral blood were done on admission using Multitest® (BD Biosciences) to look for T-cell subsets, CD19 PE-Cy7 (clone SJ25C1, BD) and CD38 APC-H7 (clone HB7, BD) to B-cell and plasmablasts discrimination. Demographic, clinical and laboratory data were collected from hospital sources.  
Results: Results: From 121 patients accessed, 76 (62,8%) had confirmed with SARS-CoV-2 infection, 47 males, median age 55.9 years (±15.9) and 81.6% with at least one comorbidity. The group was divided based on clinical factors and use of mechanical ventilation (MV) in Severe (n=41; 53,9%, no MV) and Critical (n=35; 46,1%, MV) groups. Severe illness subjects were younger (53y vs 64y), showed a lower prevalence of comorbidities (68,3% vs 97,1%), and had a lower median length of hospitalization (8.4 days, 1-35, versus 26, 6-85) (p< 0.001). The median length of MV support was 14.7 days (4-64) in Critical group. Median leucocytes in this cohort were 8930/µL, being higher in critical group (11920 vs 7120), as was neutrophil count (11400 vs 5150). On the other hand, lymphocytes were lower in Critical group (600/µL) than Severe (1050/µL), as was T-cells and subsets: CD3 (370 vs 740), CD4 (180 vs 390) and CD8 (110 vs 220), all p< 0.001. The ROC curve analysis shows that patients with CD3 more than 570 cell/µL, CD4 320/µL and CD8 120/µL had more chances to survive (p< 0.001). Mortality rate was 23.7% (n=18), and 95.1% of Severe group discharged from hospital, compared to 54.3% (n=19) from Critical group.
Conclusions: Conclusions: In our cohort the evaluation of lymphocyte count and T cell subsets can provide prognostic information for severity of COVID-19 disease and can be predictive of mortality

P47
Cytolological And Multiparametric Flow Cytometry Analysis For Diagnostic Of Breast Implant-Associated Anaplastic Large Cell Lymphoma
Ana Paula Azambuja1, Fabiola Gevert2, Raisa Mehry Oliveira3, Anne Karoline Groth4
1Universidade Federal do Paraná, Curitiba, Brazil, 2Hospital Nossa Senhora das Graças, Curitiba, Brazil, 3Hospital Nossa Senhora das Graças, Curitiba, Brazil, 4Hospital Erasto Gaertner, Curitiba, Brazil

Introduction: Introduction: Breast implant-associated anaplastic large cell lymphoma (BIA-ALCL) is a provisional entity with morphological and phenotypic characteristics indistinguishable from anaplastic large cell lymphoma, as hallmark morphology and CD30 positivity, that presents as unilateral effusion associated to silicone breast implants. As diagnostic confirmation can be difficult, multiparametric flow cytometry (MFC) looking for CD30, HLA-DR and CD25 positivity may be useful. The aim of this report is to describe cytological and flow cytometric founds of nine patients with confirmed BIA-ALCL and compare to negative patients.
Methods: Methods: From Mar/2018 and Feb/2021, 117 samples of periprosthetic fluid (PF) collection were included. All specimens were collected in dry tubes and sent immediately to the lab to quantification and characterization of pathologic CD30 cells.
Sample were stained with CD4-V450, CD45-V500c, HLA-DR-FITC, CD30-PE, CD15 PercpCy5.5, CD19-PE-Cy7, CD14+CD3-APC and CD8-APCH7. Positive cases were confirmed with HLA-DR-V450, CD45-V500c, CD45RO-FITC, CD25-PE, CD5-PerCPcy5.5, CD2-PE-Cy7, CD14-APC and CD43-APCH7. A cytospin was prepared and cored with Wright-Giemsa staining for morphological evaluation.
Results: Results: 117 PF collection from 109 patients (seven bilateral and one repeated) were analyzed in 36 months. There were nine positive cases (8.25% of patients); in one of them, the first sample was considered unavailable, so MFC sensitivity was 90.0% and specificity 99.0% in this cohort. From 108 negative samples there were twelve (10.25%) unavailable, eight (6.8%) blood contaminated, 18 (15.3%) neutrophilic exudate and 70 (59.8%) transudate predominance of mature lymphocytes.
Cytological examination of all nine positive cases revealed large, anaplastic cells with pleomorphic nuclei, prominent nucleoli, and moderate basophilic cytoplasm with frequent vacuoles. MFC showed large tumor cells (increased FSC/SSC scatter) with bright expression of CD30, CD25 and HLA-DR, CD45dim and absence of monocytic, neutrophils and B cell antigens (CD14, CD15, CD19). All had absence of CD3, seven cases had CD4 and CD2 expression, one had weak CD8, one had CD5 dim.
Conclusions: Conclusion: We describe findings of nine BIA-ALCL cases with overexpression of CD30, HLA-DR and CD25 and dim expression for T cell markers in tumor cells compared with normal/reactive samples.
These findings highlight the utility of cytologic evaluation and multiparametric flow cytometry immunophenotyping in BIA-ALCL diagnostic workup.

P48
The AbЕRrant AntigЕNЕS In The Acute LЕUcЕMiЕS.
Hanaa Bencharef
Chu ibn rochd casablanca , Casablanca, ON, Morocco

Introduction:  The еvaluation of  thе prеsеncе of blasts with an abеrrant phеnotypе, will bе onе of thе important еlеmеnts takеn into account in thе dеvеlopmеnt of futurе thеrapеutic protocols, as wеll as monitoring for rеsidual disеasе aftеr trеatmеnt. In the present study, wе rеport thе intеrеst of flow cytometry in thе diagnosis of acutе lеukеmia еxprеssing abеrrant markеrs.
Methods: A descriptive, analytical study carried out by the Laboratory of Hematology  CASABLANCA, collating, patients who present an acute lеkemia , over a period of  4-years from January 1, 2017. The immunophenotypic diagnosis is carried out on Bеckman Coultеr type Navios flow cytometer using 6  colors. The first panel of monoclonal antibodies used: is the orientation panel , used to assign cell proliferation lymphoid  or myeloid. Thе second panel: idеntify thе stage оf differentiation, prognosis оf aberrant characteristics or phenotypеs. CD45 is used to differentiate celular populations, it allows to demarcate celular contingents and target blasts. The data collected from blast cells as well as their immunophenotypic profile was carried out by computerized software KaliSil® from  our hospital , then entered and analysed by the software Excеl. 2013.
Results: 648 casеs  were analyzed  . Thе immunophеnotypic profilе of our study showеd 57% of casеs of AML and 43 % of ALL , and only 3 casеs of acutе mixеd lеukеmia. In addition, in 17% of casеs, wе dеtеctеd a co-еxprеssion of abеrrant markеrs, of which 68% of casеs wеrе associatеd with AML and 32% of casеs associatеd with ALL, abеrrant phеnoyps arе found in thе majority of casеs in AML. Thе CD7 abеrrant markеr was most еxprеssеd in AML (Table 1). Whilе for ALL, thе most common abеrrant markеr is CD13.
Conclusions: Thе rеsults of our study arе consistеnt with thosе of thе litеraturе. Flow cytomеtry thеrеforе makеs it possiblе to dеtеrminate of abеrrant Ag, it can bе usеd to follow thе еvolution of patiеnts undеr trеatmеnt, and study thе rеsidual disеasе. our study needs to be complemented by further studies which will aim to correlate these aberrant Ag with prognosis аnd therapeutic response. As well as, expanding this study at the national level, to confirm this profile and monitor the evolution of the incidence of aberrant phenotypes.

P49
Evaluation Of Minimal Residual Disease By Flow Cytometry  (Experience Of Hematology Laboratory Of Casablanca) 
Hanaa Bencharef
Chu ibn rochd casablanca , Casablanca, ON, Morocco

Introduction: The flow cytometry is an increasingly and useful technique that can be applied to discriminate of phenotypically abnormal PCs from their normal counterparts . In our context, it is currently under development in our laboratory for the post-treatment evaluation of malignant hemopathies .In the present paper, we report the experience of the casablanca hematology laboratory in the evaluation of residual disease by flow cytometry in patients followed for multiple myeloma , in order to,  validate the protocol of plasma cell immunophenotyping by this technique.
Methods: A prospective analytical study carried out in the hematology laboratory, at the Ibn Rochd university hospital in Casablanca, over a period of 8 months from September 2019. The study involved all the patients suspected MM eligible for transplant, and whose diagnosis was retained according to the diagnostic criteria of IMWG 2014.The immunophenotypic analysis was done using a Navios cytometer (Beckman Coulter), in a 6-color and 2-laser configuration. we have adopted a theoretical sensitivity of the order of 10-4 in order to interpret our results.
Results: 20 patients diagnosed with MM  were collected. The median age at diagnosis was 54 years , The immunophenotypic profiles of the patients are heterogeneous, but the most found profile was : CD38 +, CD138 +, low CD45, CD19-, CD56 +, CD20-, CD33- (Table I). The most widely chemotherapy regimen was CTD in 75% of cases. Only 5 patients were evaluated after induction chemotherapy and we found failure in three patients with positive MRD, PR in one patient and CR in one patient with negative MRD, 5 patients died, 2 of whom were diagnosed with malignant hypercalcemia and 3 during treatment; by septic shock in 1 patient and by pulmonary embolism in 2 patients, the other patients are still being treated.
Conclusions: The preliminary results allowed us at this stage to validate the flow cytometry plasma cell immunophenotyping technique, and to prove its feasibility in our context. Our outlook is for this method to become protocol-based and for it to be integrated into the management of our patients followed for MM. 

Best Student Presentation Award Finalist
P50
15 Antibody 10-Color Flow Cytometry Panel &Ndash; A Screening Tube For Hematologic Malignancies
Laiz Bento, Flavia Sousa, Nadila Millan, Priscila Miyamoto, Marilia Passaro, Andressa Vaz, Daniela Schimidell, Elizabeth Souto, Nydia Bacal
Hospital Israelita Albert Einstein, São Paulo, Brazil

Introduction: Clinical flow cytometry laboratory constantly needs develop technical procedures with improvement in quality and cost reduction. We previously developed 10-color tube/14 fluorochromes (AcMo) as a screening tube (ST) for hematologic malignancies. However, this combination showed some limitations, like no plasma cells detection and, in some cases, low performance. Recently, we developed a new screening tube (NST) with 10-color/15 AcMo. We demonstrate the improvement of this tube for diagnosis of hematologic malignancies and cost comparison with the panel ST used in our routine. 
Methods: -NST AcMo: CD4+KappaFITC/CD8+LambdaPE/CD3+CD14 ECD/CD5+CD33PC5.5/CD20+CD56PC7/CD34APC/CD19A700/CD10A750/CD38PB/CD45KO. -ST AcMo: CD4+KappaFITC/CD8+LambdaPE/CD3+CD14ECD/CD33PC5.5/CD20+CD56PC7/CD34APC/CD19A700/CD10A750/CD5PB/CD45KO and a screening panel of 5-color used to identify plasma cell, composed by CD38FITC/CD56PE/CD19PC7/CD45V450/CD138V500. We compared the NST vs ST plus panel of 5-color in 8 bone marrow (BM), 2 peripheral blood (PB), 1 cerebrospinal fluid (CSF) and 1 lymph node (LN). We used Navios Flow Cytometer and analyzed by Kaluza. Cost of reagents, time and cost of technical/medical staff were evaluated.
Results: NST was able to identify normal and malignant populations even in small cell samples. NST has great applicability in suspect of lymphoproliferative disorders and myelodysplastic syndrome (analysis of Ogata Score and aberrant antigen expressions). It was possible to identify plasma cells as screening, and detect abnormal expression of CD56, CD19 and CD45. The cost of NST is 40% lower than ST and the time of technical procedure had 10 minutes of reduction.
Conclusions: ST combination showed problems in CD5PB expression, dim or negative, when in fact was positive, and in Kappa/Lambda separation even after wash/ incubation. These problems were solved with the NST using CD5PC5.5 and Kappa/ Lambda (goat). It reduces cost and time because, the repetition to confirm Kappa, Lambda and CD5 is not necessary anymore. Furthermore, with the inclusion of CD38 in NST,it is not necessary stain the panel of 5-color as a screening of plasma cells; it bringing cost reduces. Using the NST, our service has already prepared 801 BM, 300 PB, 15 CSF, 33 pleural fluids, 9 breast seromas and 133 LN. All cases have been show good performance in separation between normal and abnormal. NST is an efficient solution for screening hematological malignancies with quality and standardization with lower cost.

P51
Flow Cytometry Assessment Of Cd34+ Viability In Thawed Units.
Regielly Cognialli 1, Diesica Ferreira 2, João S Farias3, Nicole Winter4, Diana Kluck5, Raysa Rodrigues6, Aline Hey7, Vanessa Kukla8, Giulia Campos9, Míriam Beltrame10
1Hospital de Clínicas - UFPR, Curitiba, Brazil, 2Hospital de Clínicas - UFPR, Curitiba, Brazil, 3Hospital Erasto Gaertner, Curitiba, Brazil, 4Hospital Erasto Gaertner, Curitiba, Brazil, 5Mantis, Curitiba, Brazil, 6Mantis, Curitiba, Brazil, 7Mantis, Curitiba, Brazil, 8Hospital de Clínicas - UFPR, Curitiba, Brazil, 9Hospital Erasto Gaertner, Curitiba, Brazil, 10Hospital Erasto Gaertner, Curitiba, Brazil

Introduction: Background: The number of CD34+ cells infused into patients at the time of autologous or allogeneic transplantation is a clinically important variable. In peripheral blood (PB) stem cell transplantation, the number of CD34+ cells infused is considered a predictor of haematopoietic engraftment. However, the currently accepted minimal threshold of CD34+ cells/kg was determined by counting CD34+ cells before freezing, and the loss of viable CD34+ cells during freezing, cryopreservation or thawing prior to reinfusion has not been assessed.    
Methods: Materials and methods: PB were collected using acid citrate dextrose solution A (ACD) as an anticoagulant. Cryopreservation is the only available method for the long-time maintenance of blood cells. Retrospective analysis of CD34 and viability. We analyzed by flow cytometry 325 samples from 2017 to 2020. Leukocyte, proportions of CD34+ cells enumeration and viability (7AAD) were assessed from the frozen-thawed samples. The flow cytometric was based on The International Society for Hematotherapy and Graft Engineering (ISHAGE) recommendation is to enumerate CD34-positive events and use 7-actinomycin D as a viability marker. 
Results: Results: The mean content of the divided apheresis products was 105,8 ± 99,4 103//ul leukocytes, 58% ± 17,9% viable leukocytes, 1,3 ± 1,7 103/ul CD34+ cells and 84,1% ± 9,8% viable CD34+ cells. 
Conclusions: Conclusions:  In our study viable CD34 cell assays are the standard quality measure to assess the impact of storage and cryopreservation and an adaptation of both the acquisition setting and the gating strategy is needed in order to detect viable and non-viable CD34+ cells in thawed units.

P52
Limphocytes In Sars-Cov-2 Infection; Alterations In The Leukocyte Differential.
MERITXELL DEULOFEU, MARIONA HERNÁNDEZ, MAITE QUINTANA, ANNA MARULL, JAVIER NIETO-MORAGAS, ORLANDO JIMÉNEZ, MAITE SERRANDO
HOSPITAL JOSEP TRUETA, GIRONA, Spain

Introduction: COVID-19 pandemic represents a health crisis worldwide due to its high spread and icomplications that can lead to the death. Different health professionals have been involved in diagnoses  being quickness the cornerstone in these processes. One of the most relevant hospital departments are laboratories providing reverse transcription polymerase chain reaction (RT-PCR), antigen and serology results. Changes in the leukocyte differential are common in infections. Eventhough viral etiology tends to produce lymphocytosis; recent publications concluded that in the SARS-CoV-2 patients lymphopenia is observed. The aim of this descriptive study is to relate changes in the leukocyte differential to an alteration in the average of the normal lymphocyte populations detected by flow cytometry (FC).  
Methods:  We included 87 patients with SARS-Cov-2 positive results. We performed cell blood count (CBC), leucocyte differential (diff) and immunophenotype including CD45, CD3, CD19, CD4, CD8 antibodies. We have compared absolute numbers of leukocytes and leukocyte populations (mean-SD, median 25-75 percentiles). CBC and diff were performed by Sysmex XN analyzer (Japan) and FC by FC 500 Beckman-Coulter (US). We divided patients in two groups: first using CD45-CD3-CD19 and second using CD3-CD4-CD8. In the last group, we compared CD4/CD8 ratio results with those expected in non-pathological conditions (>2.0).
Results: In our study 32.2% were women and 67.8% men. The global average age was 61.5-years-old, being 66.3-years-old for women and 59.2-years-old for men. Mortality observed was 8.05%. Mean average results obtained were: leukocytes 7.45x109/L±4.01 (4-30); neutrophils 5.65x109/L±3.68 (1-20); lymphocytes 1.17x109/L±0.68 (1-5); monocytes 0.55x109/L±0.47 (0.5-1) and eosinophils 0.06x109/L±0.12 (0.5-1). The majority of the cases showed a predominance of CD3 (mean 58.83%±16.99) as expected in non-pathological condition as well. Results for non-B non-T lymphocyte population was 24.81%±13.25, probably related to CD56 marker and higher than observed in non-reactive patients (>15%). Mean results for the CD4/CD8 ration in our study were 2.47±1.62 in accordance with normal ratio. Results for the median and 25-75 percentiles are in tables 1 and 2.
Conclusions: Morbidity and mortality of COVID-19 infection has been related to coagulopathy and severe inflammatory response of the immune system. We observed lymphopenia instead of lymphocytosis in the cases studied but with non-pathological distribution of the lymphocyte populations. Moreover, CD4/CD8 ratio is within normal range so there is no immunodeficiency following WHO criteria

P53
Bone Marrow T‑Cell Percentage: A Novel Prognostic Indicator In Acute Myeloid Leukemia
nahla ibrahim
KAMC

Introduction: Acute myeloid leukemia (AML) is an aggressive
malignancy for which overall disease-free survival is less
than 50%. Manipulation of the immune system is an interesting
and promising therapy for AML patients. We aimed
to characterize the immune system of AML patients, highlighting
the clinical relevance of total bone marrow (BM)
lymphocytes and subpopulations.
Methods: Sixty-six new AML cases
diagnosed according to WHO criteria from King Abdullah
Medical City, KSA . Analysis of BM lymphocytes and subpopulations
was done by flowcytometry.
Results: Significantly, high percentages
of BM lymphocytes, T cells, and natural killer (NK)
cells were detected in the group that achieved complete
remission (P values = 0.004, < 0.001, and < 0.001, respectively).
Overall survival (OS) was significantly prolonged
in patients with high BM lymphocytes and T cells (P values
= 0.047 and P 0.002, respectively). Multivariate analysis
indicated that BM T-cell percentage and cytogenetics
were independent prognostic factors predictive of OS (HR
4.7, P value = 0.011).
Conclusions: BM T-cell percentage constitutes
a novel host factor that can be used in combination with cytogenetics to better predict OS. Large-scale multicenter         
studies are recommended to clarify its role as a predictor of
OS and leukemia-free survival.

P54
Double Hematological And Solid Malignancy Diagnosed From Bone Marrow Studies:Casereport,
nahla ibrahim
KAMC, Makkah, Saudi Arabia

Introduction: Double primary malignancies could be divided into synchronous and metachronous depending on the interval between tumor diagnoses
Methods: 63 years male patient presented to KAMC,KSA 2018
Results: The liver,kidney were within normal range,patient negative for HCV, HBV, HIV, and H1N1.Peripheral blood show normochromic anemia,White blood cells showed leukopenia , with sever neutropenia.Marked thrombocytopenia.No blasts were seen.BM sample was particulate.Megakaryocytes were adequate in number with few hypolobated forms denoting dysmegakaryopoietic changes.Granulopoiesis was adequate but revealed mainly early granulocytic precursors and some dysplastic changes. Active erythropoiesis was noticed with dyserythropoietic changes in form of binucleated and megaloblastoid changes.blasts were around 24-30%,they were variable in size with large nucleus and multiple nucleoli,more than 5% of the blasts were positive for myeloperoxidase by immunohistochemistry denoting their myeloid nature.The BM biopsy revealed 50% cellularity and interstitial involvement by blasts in the background of trilineage hematopoiesis.Some dysplastic megakaryocytes were seen.Cytogenetic results revealed (46,XY[20]).FISH did not reveal any clonal abnormalities.Molecular studies revealed negative FLT3,NPM,BCR-ABL,and PML-RARA.Flowcytometry analysis dim CD45 positive cells in blast gated area expresse the myeloid markers:cytoplasmic MPO,CD13,CD33,CD15,andCD34 andCD117 and the non-lineage specific marker HLADR,The blasts were negative forT& B markers.The patient  diagnosis was acute myeloid leukemia with dysplastic changes,according to the WHO criteria it is AML-M2 not otherwise specified.Being an old patient with generally poor condition,the treatment owas Azacytidine to reduce the blast count.On May 2019, the serum ferritin level was very high; 2382ng/ml (RR 8-388 ng/ml) and the prolactin level was elevated; 22(0-0.1 ng/ml) On the next follow up the cytopenia was very obvious,another BM study done.The positive laboratory investigations were very high serum Ferritin level; 8109.8 ng/mL,elevated CEA;48.59 ng/dL(RR0-5)andCA19-9; 270.08 U/mL(R.R 0-30.9).In contrast to the PB examination at presentation,the myeloblasts were seen in the PB;14% and in the BM as well.The examination of the touch print& BM biopsy revealed sheets of malignant non-hematopoietic cell.Immunohistochemistry studies reveals negativeCD45,CD34,prostatic specific antigen (PSA) and anti-IL-7receptors (CDX2),the malignant non hematopoietic sheets showed positive reaction for cytokeratin 7,pancytokeratin (CKAE1AE3) and Thyroid transcription factor (TTF).The patient received palliative therapy and died two months later 
Conclusions: After the recent diagnostic and staging modalities as well as progress in the management of common cancer,the possibility of the existence of a second primary malignancies should always be considered whenever a novel laboratory finding emerges that could be useful for the early detection of associated tumors

Best Student Presentation Award Finalist
P55
Flow Cytometric Assessment Of Surface Neutrophilic Myeloperoxidase; A Risk Prediction Tool For Covid-19-Associated Coagulopathies
Elham Jamali1,2, Ali Arabi Monfared2, Ehsan Sarraf Kazerooni1,2, Parisa Tandel1, Reza Ranjbaran1
1Diagnostic Laboratory Sciences and Technology Research Center, School of Paramedical Sciences, Shiraz University of Medical Sciences, Shiraz, Iran., Shiraz, Iran, 2Peyvand Pathobiology and Genetic Laboratory, Shiraz, Iran, Shiraz, Iran

Introduction: Inflammatory response-induced coagulopathy is a common complication associated with severe form of COVID-19 infection and it is estimated that more than 70% of patients who died from COVID-19 were also suffering from disseminated intravascular coagulation (DIC). Furthermore, several compelling evidences suggest that neutrophil extracellular trap (NET) formation plays a role in triggering the immunothrombosis process in these cases. In the current study, we evaluated the surface neutrophilic MPO as NETosis biomarker for predicting the risk of covid-19-associated coagulopathies.We hypothesized that the release of MPO into plasma during NETosis could possibility lead to a decline in the cellular level of this enzyme.
Methods: The COVID-19 infection was assessed by real-time PCR while D-dimer level, a sensitive marker of thrombosis, was measured by ELISA method. Based on these two tests, patients were categorized into four different groups including COVID-19 + with elevated D-dimer, COVID-19+ with normal level of D-dimer, COVID-19 negative cases with normal D-dimer level and a group of patients with high level of D-dimer and negative for COVID-19. Using flow cytometry method, the peripheral blood leukocytes were stained with FITC-conjugated anti human myeloperoxidase and the level of neutrophilic surface MPO was compared between the studied groups. 
Results: Our findings revealed that surface expression of neutrophilic MPO significantly decreased in COVID-19 positive cases compared to COVID-19 negative groups. Moreover, results demonstrated a significant inverse correlation between the level of D-dimer and MPO expression. Consequently, MPO expression on the surface of neutrophils indicated a noticeable decrease in covid-19 patients with higher level of D-dimer when compared to those with normal D-dimer levels.
Conclusions: The proposed method can be considered as a valuable tool in predicting the risk of hypercoagulopathy in COVID-19 patients through evaluation of MPO expression level on neutrophils’ surface. Our findings also highlighted the important role of NET formation in triggering the coagulopathies in COVID-19 patients.

P56
Specific Features Of T-Cellular Immunity In Chronic Lymphocytic Leukemia.
Evgeniy Pochtar1, Svetlana Lugovskaya1, Elena Naumova1, Elena Dmitrieva2, Eugene Nikitin2
1Russian Medical Academy of professional , Moscow, Russia, 2Russian Medical Academy of professional , Moscow, Russia, 3Russian Medical Academy of professional , Moscow, Russia, 4SP Botkin Municipal Clinical Hospital, Moscow, Russia, 5SP Botkin Municipal Clinical Hospital, Moscow, Russia

Introduction: Infection complications are the main cause of mortality in chronic lymphocytic leukemia (CLL), and immunological dysfunction is one of the leading factors in their occurrence. The aim of this work is to assess the features of the subpopulation composition of T-lymphocytes (T-helpers (Th), cytotoxic T-lymphocytes (Tcyt), T regulatory cells (Treg), T-NK cells, naive Th, Th-memory, activated T-lymphocytes, TCRγδ cells) in peripheral blood of patients with newly diagnosed chronic lymphocytic leukemia (CLL) and receiving ibrutinib therapy.

Methods: Hematological and immunophenotypic studies have been performed in 172 patients (30 patients with previously untreated CLL, 122 patients on ibrutinib therapy and 20 healthy donors). The subpopulation composition of T lymphocytes (Th, Tcyt, Treg, T-NK, naive T-helpers, memory T-helpers, TCRγδ cells, activated T-lymphocytes) was assessed on a flow cytometer (FACSCanto II (BD)) using a panel of monoclonal antibodies.
Results: Compared to controls all CLL samples were found to have higher the absolute number of T-lymphocytes (p˂0,0001) by T-helpers (р=0,0044) (especially of memory T-cells (р=0,023)), cytotoxic T-cells (р˂0,0001), regulatory T-cells (р˂0,0001), TCRγδ T-cells(р=0,0128), activated T-lymphocytes for both the CD3+CD25+ population (early activation) (p˂0.0001) and CD3+HLA-DR+ population (late activation) (p˂0.0001) in previously untreated CLL patients (I group). In ibrutinib-treated patients (group II) the absolute number of both T-lymphocytes (p=0,0006) and the absolute number of subpopulations of T-NK cells (p=0,0007) decreases compared to group I patients. At the same time, the values of the above subpopulations don't reliably differ from those of donors. Also, in group II patients have been marked by significant decrease of T-lymphocytes with markers of early and late activation as compared to group I (p˂0,0001, but the values of these populations remain significantly higher than those of donors (p=0,0007- for CD3+CD25+-cells and p=0,0079 - CD3+HLA-DR+-cells).
Conclusions: CLL patients have been found to have quantitative and functional changes in the subpopulations of T-lymphocytes, indicating dysregulation of the immune response, and a high risk of developing infections. Monitoring of immunological parameters for ibrutinib therapy make possible to estimate impact of ibrutinib on the adaptive anti-CLL immune response.

P57
Single Flow Cytometry Panel For Malignant Plasma Cell Immunophenotyping Used At Diagnosis And Follow Up
Katarina Reberšek1, Tadej Furlan1, Samo Zver1,2, Helena Podgornik1,3
1University Medical Center, Ljubljana, Slovenia, 2Faculty of Medicine, Ljubljana, Slovenia, 3Faculty of Pharmacy, Ljubljana, Slovenia

Introduction: Multiple myeloma is a neoplastic disorder characterized by malignant plasma cells in bone marrow. Response criteria are based on monoclonal protein and bone marrow cytomorphological evaluation. These parameters are not sensitive enough to detect minimal residual disease. Both normal and malignant plasma cell populations can be identified using multiparameter flow cytometry at diagnosis and at follow-up.   Noteworthy, no single phenotypic marker is able to distinguish between normal and malignant plasma cells and the optimal approach is to use the exact same panel for pre and post treatment samples in order not to miss clonal evolution. We evaluated our 10-colour panel for plasma cell immunophenotyping at diagnosis and at follow up.
Methods: Bone marrow samples were lysed using ammonium chloride, washed with PBS, and incubated with monoclonal antibodies (CD38-FITC / CD138-PE / CD28-ECD / CD117- PerCP-Cy5.5/ CD200-PC7 / CD81-APC / CD56-A700 / CD19-A750 / CD27-V450 / CD45-KO). Overall, 117 bone marrow samples of patients were analyzed that were part of the routine diagnostic workup.
Results: Diagnostic sensitivity and specificity were assessed using data of bone marrow cytomorphological evaluation, histology, and cytogenetics. In our cohort, 76% (89/117) of patients were diagnosed with plasma cell dyscrasia (PCD) and 24% (28/117) were negative for PCD. Diagnostic sensitivity and specificity were 97% and 100%, respectively. Results of flow cytometry were false negative in 3 patients, i.e. 2 patients were diagnosed with solitary plasmacytoma. In all 3 patients bone marrow biopsy reported PCD with 1-5% infiltration. However, bone marrow cytomorphologic evaluations were negative for plasma cell infiltration, and no cytogenetic abnormalities were found in two patients, and there were not enough plasma cells to perform cytogenetics in one patient. In flow cytometry false negative samples, low plasma cell infiltration was obtained (range 0,2-0,8%), suggesting that hemodilution may be the key factor influencing the results.
MRD assessment using the same panel was verified through 8 samples of external quality assessment service UK Neqas. There were no significant differences between results assessed in our laboratory and other laboratories. Furthermore, statistical analysis confirmed that our results have no bias.
 
Conclusions: Our 10-colour panel has high diagnostic sensitivity and specificity in detecting malignant plasma cells and it can be used at diagnosis and at follow up.

P58
Introduction Of A Flow Cytometry Assay For The Differentiation Between Reactive Monocytosis And Chronic Myelomonocytic Leukemia At Newcastle Hospitals
Stephanie Shail, Helen Watson
Newcastle Hospitals, Newcastle upon Tyne, United Kingdom

Introduction: It is often challenging to distinguish between reactive monocytosis and chronic myelomonocytic leukemia (CMML). Flow cytometry has been identified as a potential diagnostic tool in these cases; particularly the skewing of monocyte subsets in peripheral blood (PB) to more than 94% classical monocytes (CD14+/CD16 negative), and aberrant CD56 expression on monocytes in both PB and bone marrow (BM). Introduction of a monocyte assay at the flow cytometry laboratory within Northern England Haemato-Oncology Diagnostic Service will enable timelier, accurate diagnosis for patients with CMML or reactive monocytosis within the region. Currently, the confirmation of a reactive state can only be made after significant delay, and repeated clinic visits and blood sampling which could be avoided if confirmed by flow cytometry. 
Methods: The assay was set up using the following antibodies; CD45, CD4, CD14, CD16 and CD56. Technical validation of the assay was performed as per the BSH Guidelines on the use of multicolour flow cytometry in the diagnosis of haematological neoplasms, including assaying a number of normal samples, samples from patients with reactive monocytosis, and patients with CMML.
Results: The four normal samples, all with normal full blood count indices including monocyte count showed expected results. The monocytes were CD56 negative on all normal samples tested, one normal sample showed increased classical monocytes of 94.8%, however the assay is not designed to be used in patients without monocytosis. Nine patients with monocytosis were used as part of the validation; these were mainly reactive. Two of these patients had increased classical monocytes and therefore were thought to have underlying CMML. As expected in reactive monocytosis there was less than 94% classical monocytes. CD56 expression was variable, consistent with the understanding that CD56 may be expressed in reactive monocytosis. A patient with known CMML was also assayed; this showed 96.8% classical monocytes and variable CD56 expression, flow data from this patient shown in Figure 1.
Conclusions: We have demonstrated that PB flow cytometry is a useful diagnostic tool to distinguish between reactive monocytosis and CMML, helping to identify those patients where further assessment may be appropriate. Further work is required to demonstrate the utility of this assay in BM samples.

P59
Review Of The Impact Of Introduction Of Cd200 To Chronic Lymphoproliferative Panel At Northern England Haemat-Oncology Diagnostic Service
Stephanie Shail, Helen Watson
Newcastle Hospitals, Newcastle upon Tyne, United Kingdom

Introduction: Numerous studies have demonstrated the use of CD200 as a distinguishing marker between Chronic Lymphocytic Leukaemia (CLL) and Mantle Cell Lymphoma (MCL), with CD200 ubiquitously expressed in CLL. In MCL CD200 is usually negative, with dim positivity in up to 10% of cases. Following a successful validation project, CD200 was added to the Chronic Lymphoproliferative Panel at Northern England Haemato-oncology Diagnostic Service in May 2020. During this initial validation we demonstrated the utility of CD200 as a distinguishing marker between CLL and MCL, potentially mitigating the need for additional genetic testing in some cases. The differing CD200 expression in the two conditions is shown in Figure 1. This audit assessed 100 patients who were tested following the introduction of CD200.
Methods: The flow cytometry results of 100 patients were compiled along with records of CCND1 (t(11;14)) fluorescent in-situ hybridisation (FISH) analysis. The diagnosis was obtained from Haemosys within the final integrated report. A CLL score was calculated for each patient with a CD5+ clonal B-cell population according to the Matutes scoring system.
Results: 46 out of 47 CLL samples were positive for CD200, with weakly positive expression in one case. This included one case of atypical CLL. Three of the CLL cases had Matutes scores of three, therefore were tested for evidence of a t(11;14) translocation, characteristic of MCL, and all were negative, consistent with their classification as CLL. Of the seven patients with MCL, five were CD200 negative, one had some weak CD200 expression and one case was CD200 positive. The CD200 positive MCL was an outlier case where the CCND1 translocation had been identified previously, the final diagnosis was made on that basis. The lack of CD200 was not an exclusive finding of MCL, being also observed in other B-cell proliferative disorders including, but not limited to, marginal zone lymphoma, diffuse large B-cell lymphoma and Burkitt’s lymphoma.  Furthermore, CD200 was positive in hairy cell leukemia (HCL) and HCL variant.
Conclusions: The data suggests that MCL can be effectively excluded in CD5-positive CD200-positive clonal B cell disorders. The inclusion of CD200 within our lymphoproliferative panel has meant that a diagnosis of CLL, even in cases with low Matutes score, can be made without cytogenetic analysis to rule out MCL.

Best Student Presentation Award Finalist
P60
Identification Of Peripheral Blood Cd26+ Leukemic Stem Cells Has A Potential Role In The Rapid Diagnosis Of Chronic Myeloid Leukemia
Praveen Sharma1, Neelam Varma1, Man Upesh Singh Sachdeva1, Pankaj Malhotra2
1Department of Hematology, Postgraduate Institute of Medical Education and Research, Chandigarh, India, 2Department of Internal Medicine (Clinical Hematology division), Postgraduate Institute of Medical Education and Research, , India

Introduction: INTRODUCTION: Chronic myeloid leukemia (CML) is a stem-cell (SC) neoplasm characterized by the presence of BCR-ABL1 fusion. The confirmation of diagnosis is based on detection of BCR-ABL1 through molecular or cytogenetic techniques. We performed a flowcytometric assay to identify CD26+ CML leukemic stem cells (LSCs) in a wide variety of hematological disorders (including CML and non-CML patients) and explore its value as a standalone diagnostic investigation for the diagnosis of CML.
Methods: METHODS: Patients with clinical suspicion of CML/CML on follow-up were included and peripheral (PB) and/or bone marrow (BM) samples were utilized for flowcytometric analysis. PB and/or BM of patients with diseases other than CML, patients of lymphoma staging with uninvolved bone marrow, acute leukemias, MDS/MPN, other MPNs who proved negative for BCR-ABL1 translocation by fluorescent-in-situ hybridization or reverse transcriptase polymerase chain reaction (RT-PCR) were kept as controls. Under Lyse-wash-stain-wash sequence, the sample was incubated with a pre-titrated custom-made antibody cocktail in a ‘test’ tube containing CD45, CD34, CD38 and CD26 moabs. At least 1 million events were acquired in BD FACS Canto II for the identification of CD26+ CML LSCs. Clinical data including demographic details, complete blood count and BM findings were also noted.
Results: RESULTS: A total of 91 samples (PB and/or BM) from 56 patients [suspected CML (n=26), CML on follow-up (n=12), non-CML (n=18)] were tested. CD26+ LSCs were identified in all patients with a confirmed diagnosis of CML (Median=0.07 (range 0.002-26.79)). None of the patients in the control group (non-CML) and follow-up patients with negative RT-PCR results showed the presence of CD26+ LSCs. Also, there was a strong correlation between CD26+ CML LSCs in the PB and BM (r=0.917).
Conclusions: CONCLUSION: Flow cytometric identification of CD26+ LSCs in the peripheral blood can be a cheap, rapid, robust and potential diagnostic tool for the diagnosis of CML compared to available molecular testing methods. It is independent of BCR-ABL1 transcript type and its role in residual disease monitoring is being explored.

Best Student Presentation Award Finalist
P61
Abnormalities Of T Helper 17 Cells In Acquired Aplastic Anemia And Impact Of Human Recombinant Interleukin -17A On Hematopoiesis
Vandana Sharma, Manju Namdeo, Prabin Kumar, Sandhya Sorra, Joyeeta Talukdar, Uma Kanga, Sudha Sazawal, Sujata Mohanty, D.K. Mitra, Tulika Seth
All India Institute of Medical Sciences, New Delhi, India

Introduction: Acquired aplastic anemia (aAA) is an auto immune bone marrow failure syndrome characterized by peripheral blood (PB) pancytopenia and bone marrow (BM) hypocellularity. Postulated pathology involves excessive production of T helper cells (Th) and cytokines that attack hematopoietic stem cells (HSCs). Among Th cells, Interleukin-17A (IL-17A) producing CD4+Th subset (Th17) is associated with autoimmune diseases. In this study we investigated Th17 cells, plasma level of IL-17A in PB and BM of untreated aAA patients and healthy controls by flow cytometry and ELISA respectively. Furthermore, in vitro effect of recombinant human IL-17A (rhIL-17A) cytokine was assessed on HSCs by colony forming unit (CFU) assays in methylcellulose.
Methods: Heparinized PB and BM paired samples collected from 30 untreated aAA patients; 20 PB and 10 BM samples from healthy controls. Two colour flow cytometry used for identifying Th17 cells in BD FACS CALIBUR. Plasma IL-17A quantified by ELISA. Mann-Whitney U test used for statistical analysis. For CFU assays, healthy BM mononuclear cells plated in duplicate with or without rh-IL-17A (0.1µg/ml, 0.2µg/ml, 0.5µg/ml). On day 14, colonies counted as burst forming unit-erythroid (BFU‑E), colony forming unit-granulocyte/macrophage (CFU-GM), colony forming unit- granulocyte /erythroid/macrophage/megakaryocyte (CFU-GEMM). Friedman test followed by post hoc analysis used to detect differences in CFUs. Values expressed as mean±sd. P < 0.05 considered significant.
Results: Increased percent frequency of Th17 cells in PB and BM of patients (0.62±0.16 & 0.60±0.22) than healthy controls (0.47±0.14 & 0.27±0.07) respectively. Elevated IL-17A plasma levels in PB and BM of aAA patients (135.6±7 &138.7±7.1) compared to controls (81.7±6.7 & 52.37±9.2) respectively. No difference between PB and BM of patients. Addition of rhIL-17A showed inhibitory effects on healthy CFUs compared to those without rhIL-17A.
Conclusions: The findings implicate increased Th17 cells and IL-17A cytokine in pathophysiology of aAA. Our study also provides the experimental support for this association, as the number of healthy CFUs decreased with increased doses of rhIL-17A cytokine. Thus, Th17 cells and their effector cytokine IL-17A might be attacking the hematopoietic stem cells and contribute to bone marrow failure in acquired aplastic anemia.

Best Student Presentation Award Finalist
P62
Blood Cell Microparticles In Covid-19 Patients And Healthy Donors
Olga Sirotkina1,2,3, Alexey Ermakov 4, Larisa Gaykovaya4, Tatiana Vavilova1
1Almazov National Medical Research Centre, Saint-Petersburg, Russia, 2PNPI named by BPKonstantinov of NRC Kurchatov Institute, Saint-Petersburg, Russia, 3Pavlov University, Saint-Petersburg, Russia, 4North-Western State Medical University named after IIMechnikov, Saint-Petersburg, Russia

Introduction: The monitoring of COVID-19 patients showed the development of coagulopathy. A significant increase in the number of microparticles of platelet and leukocyte origin indicates haemostasis activation and may be a predictor of adverse thrombotic events in various diseases. The inflammatory process and severe infection pathology activate the hemostasis system and have a damaging effect on the endothelium of blood vessels, so it is logical to assume the active separation of microparticles from blood cells in COVID-19 patients. The aim of this work is to analyze the number of microparticles that separate from blood cells in COVID-19 patients and in healthy donors.
Methods: The study included 40 individuals (age 21-79): 18 COVID-19 patients and 12 controls without signs of acute respiratory disease and cardiovascular or thromboembolic episodes in the history. All patients and controls underwent clinical blood tests on a 5-diff hematology analyzer. The number of microvesicles was measured by flow cytometer using an Exo-FACS reagent kit (HansaBioMed Life Sciences, Estonia) and fluorescently labeled antibodies to CD41(platelets), CD45(white blood cells), CD235a(red blood cells) (Beckman Coulter, USA).
Results: COVID-19 patients showed an increase in the relative number of neutrophils (p=0.001) compared to controls, with a simultaneous decrease in lymphocytes and eosinophils (p< 0.0001).The number of platelets in patients was also reduced compared to the controls (p=0.03).The number of exosomes was significantly higher in patients compared to controls (%positive events) - 89.2[82.3;96.4] compared to 56.2[50.5;73.5], respectively (p=0.00001) (Fig1). In terms of cell affiliation, CD41+, CD45+ and CD235a+microvesicles prevailed in patients, against those in controls (Fig2). There was no correlation between age, sex or the severity of the disease and the number of microvesicles in patients and in controls. However, a correlation was found between the number of CD45+microvesicles and the leukocyte count in patients, but not in controls. On the contrary, in the control, a correlation was found between the number of CD41+microvesicles and the number of platelets. In general, the number of CD9+exosomes and neutrophils were correlated too.
Conclusions: An increase in the number of microparticles in COVID-19 patients may explain pathophysiological mechanisms of activation of the blood coagulation system and the development of coagulopathy.The research is supported by RFBR No.20-04-00257.

P63
Downregulation Of Long Noncoding Rna Meg3 Predicts Poor Survival In Multiple Myeloma
Nashwa El-Khazragy1, Hanaa Fathey Abdelsamee2, Hagar H. Fahim3, Anwar Hatem4, Gehan Safwat3
1Department of Clinical Pathology-Hematology and AinShams Medical Research Institute (MASRI), Cairo, Egypt, 2Internal Medicine-Clinical Hematology Department, Faculty of Medicine, Ain shams University, Cairo, Egypt, 3Faculty of Biotechnology, October University for Modern Sciences and Arts, Giza, Egypt, 4Faculty of Medicine, Ain shams University, Cairo, Egypt

Introduction: Multiple Myeloma (MM) is a heterogeneous disease that is described by the accumulation and multiplication of plasma cells in bone marrow. Over the recent twenty years; developments in treatment have improve the disease outcome and make extraordinary contrasts in patient’s survival. However, over 20% of them eventually relapsed. Therefore, discovery of promising prognostic biomarkers is needed for surveillance and monitoring of therapy in MM patients. Long non-coding RNAs (lncRNAs) are a tiny non-coding RNAs comprises of 200 nucleotides or more. Aberrant expression of LncRNA has been involved in critical hemopoietic biological processes, such as cell proliferation, differentiation, metabolism and apoptosis. They transcriptionally manage basic cell forms through functioning as either oncogene or tumor suppressor genes. The Maternally expressed gene 3 (MEG3), is a noncoding lncRNA, locates on human chromosome 14q32.3 region. Downregulation of MEG3 expression has been demonstrated in variety of cancers, including chronic myeloid leukemia (CML), T-cell acute lymphoblastic leukemia, lung cancer, Hepatocellular carcinoma, prostate cancer, meningioma and neuroblastoma. Our study aimed to investigate the prognostic value of lncRNA “MEG3” in MM, and to correlate its expression with patient’s Clincopathological characteristic’s, survival profiles and response to treatment.
Methods: The expression of lnc-MEG3 was measured in 95 bone marrow samples of de novo patients with MM (prior to treatment), and 15 healthy donors by quantitative Real Time PCR (qPCR).
Results: We demonstrated a significant decrease of lnc-MEG3 expression by 5.6 folds in MM patients compared to healthy control. Patients were classified according to the median expression level into moderate/marked reduction groups. Marked reduction was significantly associated with disease at high risk and poor prognosis.The Kaplan-Meier survival analysis displayed that lower expression level of lnc-MEG3 was significantly associated with reduced overall survival (OS) and event-free survival (EFS). Moreover, Cox regression analysis revealed that lower lnc-MEG3 gene expression was independent predictor for worse survival in patients with Multiple Myeloma.
Conclusions: Our results highlights the prognostic application of lnc-MEG3 as novel promising independent biomarker in Multiple Myeloma, and it may serve as therapeutic target in future.

P64
Aberrant Cytoplasmic Cd3 Expression Is Frequently Detected In Ram-Phenotype And Cbfa2T3/Glis2 Fusion-Positive Pediatric Non-Down Syndrome Acute Megakaryoblastic Leukemia (Non-Ds-Amkl)
Ghazaleh Eskandari, Amos Gaikwad, Andrea Marcogliese, Lizmery Ferguson, Julienne Brackett, Jyotinder Punia, Tarek Elghetany, Reshma Kulkarni, Pulivarthi Rao, Jo Ringrose, Dolores Lopez-Terrada, Angshumoy Roy, Kevin Fisher, Choladda Curry
Departments of Pathology & Immunology and Pediatrics, Baylor College of Medicine & Texas Children’s Hematology-Cancer Center, Houston, TX, United States

Introduction: CBFA2T3-GLIS2
fusions, identified in pediatric non-DS-AMKL, are associated with RAM-phenotype (bright CD56, dim-to-negative CD45 and CD38, and absent HLA-DR). Rarely, cytoplasmic CD3 (cCD3) is identified in infant non-DS-AMKL, making mixed phenotype acute leukemia (MPAL) a controversial diagnostic consideration. We investigated the relationship among cCD3, RAM-phenotype, and CBFA2T3/GLIS2 fusions in pediatric non-DS-AMKL.
Methods: Clinicopathologic and laboratory data of non-DS-AMKLs diagnosed from 2014-2020 were retrospectively reviewed. cCD3 expression and RAM-phenotype were confirmed by reviewing flow cytometry data or immunohistochemical cCD3 expression status. Targeted RNA next-generation-sequencing (RNA-NGS), FISH, and statistical analyses were performed.  
Results: Results are summarized in Table 1. Twelve non-DS-AMKL patients (1 month-18 years, median 1.9 years; M:F ratio 1.4) had available RAM-phenotype status with RNA-NGS performed and/or FISH data available. RNA-NGS and/or FISH (n=11) revealed CBFA2T3/GLIS2 (n=4), KMT2A rearrangements (n=3), RBM15/MLK1 (n=1), PICALM/MLLT10 (n=1), or no fusions (n=2). All 4/4 CBFA2T3/GLIS2-positive cases and 1/7 CBFA2T3/GLIS2-negative cases exhibited RAM-phenotype (p=0.015). Aberrant cCD3 expression was identified in 4/11 cases by flow cytometry with blasts ranging from 31-49% positivity. A trending correlation between cCD3 expression and RAM-phenotype (p=0.061; 4/6 RAM-phenotype and 0/5 non-RAM-phenotype) was observed but not CBFA2T3/GLIS2 fusion status (p=0.191 ; 3/4 CBFA2T3/GLIS2-positive and 1/6 CBFA2T3/GLIS2-negative). The outcomes were poor with only 5 patients alive at follow-up including 2 without evidence of disease. Although statistical analysis was limited due to small sample size and variable treatment regimens including bone marrow transplantation in 8 patients, CBFA2T3/GLIS2-fusion-positive status, RAM-phenotype, and cCD3 expression appear to adversely affect outcomes.  
Conclusions: The dismal outcomes in CBFA2T3-GLIS2 fusion-positive and/or RAM-phenotype in non-DS-AMKLs underscore the need for risk-adaptive therapeutic intervention; the identification of RAM-phenotype should prompt an evaluation of CBFA2T3-GLIS2 fusion. Aberrant cCD3 may provide an additional clue for CBFA2T3-GLIS2 fusion and RAM-phenotype and deter an MPAL diagnosis.  

P65
Bcr-Abl1/Jak2V617FDouble Positive In Chronic Myeloid Leukemia Patients
Rim Frikha 1, Fatma Turki1, Moez Elloumi2, Hassen Kamoun1
1Department of Medical Genetics, Sfax, Tunisia, 2Department of Hematology, sfax, Tunisia

Introduction: According to the literature data, coexistence of BCR-ABL1-JAK2V617Fis the most frequent coexistence designated in chronic myeloid leukemia (CML) patients undergoing tyrosine kinase inhibitors (TKI) therapy. Therefore, issues concerning frequency and prognostic value of the combined mutations remain poorly explored. Here we describe the clinical and molecular data of patients showing the presence of BCR/ABL1 and JAK 2 mutation.
Methods: Quantitative assessment of the BCR-ABL transcript was performed using the Cepheid Xpert BCR-ABL ultra assay. Molecular response and outcome were evaluated; according to the European Leukemia Net guidelines. Genotyping of JAK2 mutation was achieved by ASO-PCR.
Results: Co-occurrence of BCR-ABL1 with JAK2V617F was detected in 5 patients among a total of 13 CML patients under TKI. According to the ELN criteria, optimal and failure were noted in 2 (40%) and 3 (60%) CML patients with double positivity, respectively.
Conclusions: Our finding suggest that CML with BCR-ABL1/JAK2V617F double positive may be a specific entity and further large-sample studies are needed to elucidate the pathogenesis, prognosis and optimal treatment.

Best Student Presentation Award Finalist
P66
Rapid Detection Of Bcr-Abl1-Like Acute Lymphoblastic Leukemia Using Machine Learning Classifiers
Dikshat Gupta1, Neelam Varma1, Shano Naseem2, Man Updesh Singh Sachdeva2, Parveen Bose2, Meenkashi Gupta2, Jogeshwar Binota2, Preeti Sonam2, Palak Rana2, Pankaj Malhotra3, Subhash Varma3, Ashish Kumar4
1Department of Hematology, Post Graduate Institute of Medical Education and Research., Chandigarh, , India, 2Department of Hematology, Post Graduate Institute of Medical Education and Research., Chandigarh, , India, 3Department of Internal Medicine, Post Graduate Institute of Medical Education and Research., Chandigarh, , India, 4International Centre for Genetic Engineering and Biotechnology (ICGEB)

Introduction: In the revised edition of WHO classification (2017) of acute lymphoblastic leukemia (ALL), a new provisional entity of “B-lymphoblastic leukaemia/lymphoma, BCR-ABL1-like” has been introduced. Worldwide, two approaches namely Gene Expression Profiling (GEP) or Next-Generation Sequencing (NGS) and TLDA (TaqMan low-density array) are used for the detection of BCR-ABL1-like cases. In India, both these techniques are not commerically available, therefore we planned to develop a logistic regression based classifier for rapid detection of BCR-ABL1-like ALL.
Methods: Flowcytometric immunophenotying (FCM-IP) and multiplex RT-PCR were performed on 551 B-lineage ALL cases to detect BCR-ABL1 chimeric fusion transcripts. Further, 12 BCR-ABL1-positive cases were subjected to transcriptome profiling using Affymetrix microarray (Gene U133 Plus Array). Lastly, we did quantification of selected 8 best genes on 551 B-ALL cases.  
Results: Incidence of BCR-ABL1 transcripts was 108/551 (19.60%) as revealed by multiplex assay. Global transcriptome profiling of 12 BCR-ABL1 RNA transcripts revealed a total of 1574 as differentially expressed (DE) genes. DE genes were further filtered through hierarchal clustering analysis and a total of 45 DE genes were identified. To identify the best classifier genes, log2 normalised expression values were analysed using penalized logistic regression. Based on previous literature and regression coefficient values, 8 genes were finally selected. 104/551 (18.87%) BCR-ABL1-negative cases were clustered together with 108 BCR-ABL1-positive cases and were categorized as BCR-ABL1-like ALL cases in the hierarchal clustering & principal component analysis (PCA). Further, we created various machine learning based classifiers using the significant predictors and finally selected a logistic regression based classifier. Lastly, we generated a dynamic nomogram for the rapid detection of BCR-ABL1-like ALL cases.
Conclusions: We have reported the incidence & devised a logistic regression based classifier, for the first time in Indian patients. Our approach is very simple, efficient, cost-effective and without any complex mathematical calculations. It is very important to detect BCR-ABL1-like ALL cases, so as to provide them the benefit of desirable TKI therapy.

P67
Genomic Classification Of Relapsed And Refractory Acute Myeloid Leukemia.
Danielle Meunier1, Iyare Izevbaye1,2
1University of Alberta, Edmonton, AB, Canada, 2Alberta Precision Laboratories, Edmonton, AB, Canada

Introduction: Acute myeloid leukemia (AML) is an aggressive malignancy with diverse genetic abnormalities that often fall outside the current World Health Organization classification system. A more comprehensive genomic classification system was recently described which categorizes AML into 11 genomic classes with distinct biological features and clinical outcomes. The purpose of this study was to characterize our local institution’s adult AML population according to the genomic classes, with an emphasis on relapsed and refractory cases.
Methods: We reviewed all cases of adult AML diagnosed by bone marrow biopsy or peripheral blood flow cytometry at our institution from 2015-2019. For those who received intensive induction chemotherapy, we compared the age, sex, blast count, and genomic classification of patients who relapsed versus those who maintained durable remissions.
Results: We identified 375 diagnostic cases AML, with treatment data available for 352. Among patients who received intensive induction chemotherapy (n=179) the median age was 56 years (range 18-87 years) with a slight male predominance (54%). A complete hematologic remission (CR) was achieved in 155 (87%) patients after one cycle of induction chemotherapy, 11 (6%) after two cycles and one (0.6%) after three cycles, while 13 (7%) patients had no documented CR. At least one relapse was documented in 49 patients, including 6 who relapsed with extramedullary disease (myeloid sarcoma). The median age and blast percentages were similar between the relapsed (59 years, 56%) and non-relapsed groups (54 years, 60%). We determined the genomic classification for 137 patients (Figure 1). The lowest rates of relapse were seen in the t(15;17), t(8;21) and CEPBABiallelic groups. Inversion 3 (inv(3)) was the most predictive of no CR, will all three cases in the series failing multiple cycles of induction. NPM1 was the most common genomic class in both the relapse and non-relapse groups (56% and 31% respectively). Rates of NPM1 with FLT3 internal tandem duplication (FLT3ITD) co-mutation were similar between these two groups. NPM1 and TP53-Aneuploidy were the most prevalent genomic classes among relapsed cases, however the time to relapse varied depending on FLT3ITD co-mutation and presence or absence of TP53 gene disruption (Table 1).
Conclusions: Genomic classification improves the prediction of relapse and refractory AML, with TP53, inv(3) and NPM1 with FLT3ITD co-mutation being important genomic markers of adverse outcomes.

P68
Next Generation Sequencing For Detection Epigenetic Regulatory Genes In Acute Myeloblastic Leukemia
Maria Molina Zayas1, Maria del Mar del Aguila Garcia2, Soledad Garcia Chileme3
1Hospital Universitario San Cecilio, Granada, Spain, 2Hospital Universitario Virgen de las Nieves, Granada, Spain, 3Hospital Universitario de Jaen, Jaen, Spain

Introduction: INTRODUCTION:
  A 71 year old male patient goes to the emergency room after presenting high fever for three days, with no additional respiratory symptoms apart from his usual dyspnea. Laboratory tests are performed: Hemoglobin  9.7 g/Dl [11-17], leukocytes: 1270 [3500-10500], neutrophils: 270 [1500-7700], platelets: 110000 / uL [120000-450000], C reactive protein: 15.95 mg / L [0.02-5]. The rest of biochemistry and coagulation levels are within normal values. A bone marrow immunophenotype is performed and the study continues.
Methods: The results of the immunophenotype were compatible with Acute Myeloblastic Leukemia (AML), myelomonocytic type (M4 according to the FAB-French-American-British nomenclature). A study of cytogenetic alterations and FLT3-ITD and NPM1 mutations is performed, analyzing fragments in the ABI3130xl analyzer, the result of which was negative. In addition, the study of mutations is extended by applying NGS (Next Generation Sequencing) techniques using a commercial panel comprising 54 genes related to myeloid neoplasms.
For sequencing the MiSeq platform was used and subsequently the analysis was performed using the Sophia DDM® application.
Results:  
GENE VARIANT PROTEIN VARIATION FREQUENCY COSMIC REFERENCE (Catalogue of somatic mutations in cancer)
IDH1 c.394C>T p.Arg 132 Cys 42,2% COSM28747
RUNX1 c.404 G>A p.Arg 135 Lys 50,3% COSM96546
SRSF2 c.284 C>T p. Pro 95 Leu 36,7% COSM146288

Conclusions: Recent studies indicate that somatic mutations in the FLT3 and NPM1 genes have prognostic relevance in patients with AML with normal cytogenetics. However, certain variants were found in other genes of interest: RUNX1 which is a transcriptional factor, SRSF2 which is a splicing gene and IDH1 that is an epigenetic regulator by DNA methylation. This isocitrate dehydrogenase gene catalyses the oxidative decarboxylation of isocitrate to alpha-ketoglutarate so its mutations block the production of alpha-ketoglutarate, a key factor for multiple proteins that regulate the epigenome. Molecules that target these IDH mutations are currently in clinical trials with early promising results for AML. In recent years, a large number of mutations in genes that regulate epigenetic control in cancer have been identified. These findings represent a great step in the follow-up and treatment of patients with AML, allowing the development of new personalized and more effective drugs for each patient.

P69
Molecular Study Of Homozygous Beta-Talasemia In Children
Maria Molina Zayas1, Soledad Garcia Chileme2, Maria del Mar del Aguila Garcia3
1Hospital Universitario San Cecilio, Granada, Spain, 2Hospital Universitario de Jaen, Jaen, Spain, 3Hospital Universitario Virgen de las Nieves, Granada, Spain

Introduction: One-year-old patient presenting irritability, paleness, yellowing of the skin and conjunctiva. In addition, he refers several visits to the pediatric emergency room due to episodes of diarrhea and fever.
Methods: Given these symptoms, a complete analysis is performed, including biochemistry and blood count. The blood test shows a decrease in hemoglobin concentration, mean corpuscular volume and mean corpuscular hemoglobin. The blood smear shows the presence of erythroblasts.
Results: Molecular genetic study is performed by hybridization with allele-specific oligonucleotides, studying 8 mutations (CD39, IVS1: 110, IVS1: 6, IVS1: 1, IVS2: 745, IVS2: 1, -87, CD6-A), which present approximately 90% beta thalassemia from the Mediterranean region. The results obtained were the following:   Mother: Heterozygous carrier for the IVS1: 110 mutation. Father: Heterozygous carrier for the IVS1: 110 mutation. Child: Homozygous affect for the IVS1: 110 mutation. The differential diagnosis includes congenital sideroblastic anemia, congenital dyserythropoietic anemias, and other situations with elevated fetal hemoglobin levels, such as juvenile myelomonocytic leukemia, bone marrow aplasia, or hereditary erythroblastopenia.
Conclusions: Both parents have an asymptomatic form of thalassemia known as beta thalassemia minor. In the case of the son, it presents the homozygous and more serious form known as beta thalassemia major, which is associated with splenomegaly and microcytic and hypochromic anemia. Onset generally occurs between 6-24 months of age.   Beta-thalassemia is inherited in an autosomal recessive manner, given that both parents are heterozygous carriers, the theoretical risk for the offspring is: 25% homozygous affect, 50% heterozygous carrier, 25% non-affected non-carrier. Genetic counseling is recommended to allow at-risk couples an informed choice among the available options, including prenatal diagnosis for future pregnancies, and thus avoid this serious disease.

P70
Relationship Between P53 Mutations And P53 Protein Expression In Chronic Lymphoid Leukemia:
Amine Moueden1, Jitka Malcikova2, Driss Benlaldj1, Reda Messaoudi1, Fatima Seghier1
1Faculty of medecine, Oran, Algeria, 2CMBGT, IHOKUniversity Hospital Brno Cernopolni , Brno, Czech Republic

Introduction: In chronic lymphoid leukemia TP53 mutations are associate with poor prognosis drug resistance and short patient survival; it is the only biological parameter that requires a change in the therapeutic strategy in CLL.  In normal cells, the concentration of P53 is generally lower than the detection threshold of immunocytochemical methods (half-life of less than 6 hours). However, mutations affecting the P53 gene can lead to the accumulation and overexpression of the P53 protein mutant. It therefore becomes detectable by immunocytochemical techniques.
Methods:  Immunocytochemical technique (ICC ) was chosen as a functional screening technique for TP53 alterations,The antibody used in our study is the anti human P53 Dako clone 318-6-11 (rabbit antibody) and visualization was done by the Dako EnVisionTM Flex system . And then detection of TP53 mutations was made by the analyzing of the whole coding region (exons 2 -11) using targeted next generation sequencing with high detection sensitivity reaching 1% allelic burden .
Results: Patient N ° 01 presents a 626_627 microdeletion at the level of exon 06 of the genomic DNA. It is responsible for a reading shift giving rise to a p.R209fs protein which presents a weak immunoreactivity revealed by ICC reactions weak. In our study, the average percentage of positive P53 cells accompanying this mutation is the lowest compared to all the mutations found (53.5%). Patient N ° 02 presents a c.313G>C mutation at the p.Gly105Arg codon. This is a mutation that is located outside the splicing sites and does not show a dominant negative effect tested on the 02 promoters WAF1 and RGC. Patient # 03 has the c.661G> T nonsense mutation at codon 221, this mutation is outside the splice sites. The other patients have a positive immunocytochimical reaction but sequencing did not find any mutations.
Conclusions: Our study found that the 03 patients with a P53 mutation present an aberrant expression of the P53 protein, but a positive immunocytochemical reaction is not always due to a direct mutation of the P53 gene but to other mechanisms of p53 stabilization.

P71
Plus One: Covid-19 Complicating A Case Of Leukemia… An Early Approach May Be The Key…
Naseelah Mussa, Ana Alves, Fátima Carriço , J. Melo Cristino
Hospital Santa Maria- Centro Hospitalar , Lisbon, Portugal

Introduction: Coronavirus disease-19 (COVID-19) is a respiratory viral disease that commonly presents with mild symptoms. It can cause serious complications such as acute respiratory disease, especially in patients with comorbidities. The data about the risk in patients with hematological malignancies is limited.
Methods: : A 42 year-old-male with Chronic Myeloid Leukemia in blastic fase, under chemotherapy after failure to respond to Tyrosine Kinase Inhibitors (TKI), was transferred to our hospital after testing positive for COVID-19 during an outbreak in a hospital ward. After few days the patient started with fever and respiratory symptoms including shortness of breath.
Results: Lung CT images revealed bilateral segmental consolidation of the inferior pulmonary lobes with air bronchograms and few ground glass opacities mainly in the right middle and inferior lobes consistent with bilateral pneumonia with acute respiratory distress syndrome (ARDS) due to the progressive hypoxia. The patient evolved with pancytopenia, sepsis, levels of Interleucine-6 of 597.8 pg/ml suggestive cytokine storm, but improved after 4 weeks of mechanical ventilation, corticosteroids, antibiotics, one dose of Tocilizumab. Sixteen weeks later the patient tested negative for COVID-19.
Conclusions: Detect and treat early CML patients with COVID-19 is crucial. Segmental consolidation and absence/few ground-glass opacities are uncommon findings on lung CT images of COVID-19 as in our case and might be complicated by ARDS. Supportive management, including Tocilizumab to control cytokine storm are of great importance. TKIs may have antiviral activity on SARS coronaviruses replication, and our patient failed to respond to it. The available data on COVID-19 in CML patients are limited. Recommendations are to manage them as healthy persons, especially if they are stable, as CML does not increase their risk of complications. Some studies showed that the risk of severe symptoms is higher in patients who failed to respond to CML treatment and that might be the case of our patient.

P72
Analysis Of Mefv Gene Mutations In Familial Mediteranean Fever Patients In Northern And Southern Provinces In Egypt.
Manal Wilson1, Safa Meshaal1, Alia El Dash1, Rabab El Hawary1
1Cairo University, Giza, Egypt, 2Cairo University, Giza, Egypt, 3Cairo University, Giza, Egypt, 4Cairo University, Giza, Egypt

Introduction: Familial Mediterranean fever (FMF) is a hereditary autoinflammatory disease which is included in the heterogeneous group of hereditary recurrent fever syndromes. It is mainly found in individuals of Mediterranean and Middle Eastern ancestry. Egypt, due to its unique location between  three continents  have always made it a coveted country by several empires. This resulted in a unique demographic and genetic distribution within the population. Recently, several studies of the MEFV gene mutation spectrum have been conducted in several districts of Egypt. However, none of these studies included a large population as the current study. Also due to the location of our hospital in the capital city, Cairo,it receives a wide variety of patients from all over Egypt.  
Methods: A retrospective analysis was made on  2000 pediatric patients aged from 5 months to 18 years who were diagnosed with FMF at  Cairo University Children’s hospital  from 2017 to 2020.  MEFV gene screening of the patients was carried out by Reverse hybridization strip assay (Vienna Lab, FMF StripAssay ). Total genomic DNA was extracted from peripheral blood samples with a DNA isolation kit. Multiplex PCR amplification was performed using biotinylated primers and the PCR products were incubated to nitrocellulose strips for reverse hybridization.
Results: Among the 2000 patients in our study, 48.5% (969/2000) of the patients had mutations in the  MEFV gene.The mutations identified were classified into four categories according to allele status.The most common pattern was heterozygous genotype in 50.1% (486/969),  compound heterozygous in 31.47% (305/969), homozygous in 14.96% (145/969), and complex genotype in 3.4% (33/969) respectively.The most common mutations according to allele frequency was M694I, followed by E148Q, V726A, M680I, M694V, A744S, P369S, Del1692, R781H and K695R respectively. Detected genetic mutations were correlated with the presenting clinical picture.
Conclusions: Familial Mediterranean fever is a highly prevelant disease in Egypt. Increasing the awarness of the clinicians of its high incidence, is important for early diagnosis and better management to avoid FMF complications, mainly amyloidosis. The technique used is a relatively simple and accurate method for diagnosis of FMF. Patients having negative results in the studied most common 12 mutations yet having criteria of FMF, should undergo sequencing of the MEFV gene.

P73
The Calreticulin Mutation Profiles Among Malaysian Patients With Philadelphia Negative Myeloproliferative Neoplasms
Razan Hayati Zulkeflee1,2, Zefarina Zulkafli1,2, Muhammad Farid Johan 1,2, Azlan Husin2,3, Md Asiful Islam1,2, Rosline Hassan1,2
1Department Haematology, School of Medical Sciences, 16150,Universiti Sains Malaysia, KUBANG KERIAN, Malaysia, 2Hospital Universiti Sains Malaysia, KUBANG KERIAN, Malaysia, 3Department Medical, School of Medical Sciences, 16150,Universiti Sains Malaysia, KUBANG KERIAN, Malaysia

Introduction: Myeloproliferative neoplasms (MPNs) are clonal over-proliferation of hematopoietic disorders  lead to excessive myeloid differentiation consist of Chronic myeloid leukaemia (BCR-ABL1 positive), Polycythaemia Vera (PV), Essential thrombocythemia (ET), Primary myelofibrosis (PMF), Chronic neutrophilic leukaemia, Chronic Eosinophilic leukaemia, NOS and Myeloproliferative neoplasm, unclassifiable. The most common driver gene mutations identify are JAK2V617F, Calreticulin (CALR) mutations and MPL (Myeloproliferative Leukemia).The calreticulin (CALR) gene/protein is located on chromosome 19p13 in humans and contains 9 exons mainly involved in calcium homeostasis, chaperone in interacting with MPL and JAK2 /STAT signalling. The mutations consist of type 1 and type 2. It has provided a new molecular diagnostic marker particularly in ET and PMF
Methods: One hundred fifty nine MPN patients were recruited from haematology clinic from January 2017 until December 2020. The diagnoses was made based on WHO classification (2016) criteria. Laboratory profile were retrieved. CALR mutations were performed by Allele specific oligonucleotide polymerase chain reaction.
Results: Total of (47.8 %, 25.8 % and 26.4 %) were diagnosed as PV, ET and PMF respectively. CALR mutations were found in 5.66% (9/159). CALR mutations were detected in 14.3% (6/42) in PMF patients, 7.31% (3/41) ET and none in PV. The type 1 mutation was the most common (77.8%) type among all patients. Those with CALR mutations were all negative for JAK2 and MPL mutations. LDH showed significantly high (p< 0.05) in PMF with CALR mutations compared to ET. Only one PMF patient with CALR mutations transformed to acute myeloid Leukemia. 
Conclusions: In conclusion, CALR is exclusively detected in JAK2 negative PMF and ET. Thus, CALR is a potentially valuable marker to stratified JAK2-negative MPN patients.

Best Student Presentation Award Finalist
P74
Modular System For Neutrophil Recognition And Myelodysplastic Syndrome Diagnosis
Andrea Acevedo Lipes1,2, Anna Merino1, Laura Boldú1, Ángel Molina1, Kevin Barrera2, José Rodellar2
1CORE Laboratory. Biochemistry and Molecular Genetics Department. Biomedical Diagnostic Center. Hospital Clinic, Barcelona, Spain, 2Technical University of Catalonia, Barcelona, Spain

Introduction: The objective of this work is to design a modular system based in Convolutional Neural Networks (CNNs) trained from scratch and able to automatically recognize normal and dysplastic neutrophils among all the nucleated cell (NC) images of the smear and platelets, as a support for myelodysplastic syndromes (MDS) detection.
Methods: We developed the modular system presented in Figure 1. The system consists of two CNNs: Model 1 and Model 2 (DysplasiaNet). For Model 1, the input is a set of all types of NC circulating in blood and the output is the classification in eight groups of blood cells (neutrophils, eosinophils, basophils, lymphocytes, monocytes, immature granulocytes, erythroblasts and platelets). Model 1 was trained with the dataset implemented in (Acevedo et al. 2019). The set containing neutrophil images was modified replacing 1,500 normal neutrophils by images corresponding to hypogranulated neutrophils. It was tested with images not used previously, obtaining an accuracy of 95.4 % in the automatic recognition of the eight groups.  In a second step, the output of Model 1 was connected to the input of DysplasiaNet, which was previously trained with a dataset of 13,362 images of dysplastic (hypogranulated) and normal neutrophils to detect MDS.  
Results: We implemented a proof of concept using NC images obtained from 66 smears belonging to 32 MDS patients and 70 smears from 40 healthy controls. Figure 2 presents the confusion matrix of the results with a total accuracy of 94.8 %. Using this setup, the pathologist does not require to select the neutrophils to feed DysplasiaNet, since all the NC images are given to the integrated system.
Conclusions: A modular sequential system was designed for the automatic selection of neutrophils in a first step and the further recognition of those showing dysplasia in a second, with high accuracy values (94.8 %). This is encouraging for further works where other models could be integrated to detect malignancies, such as leukemia and lymphoma.   References: Acevedo, Andrea, Santiago Alférez, Anna Merino, Laura Puigví, and José Rodellar. 2019. “Recognition of Peripheral Blood Cell Images Using Convolutional Neural Networks”. Computer Methods and Programs in Biomedicine 180: 105020. https://doi.org/10.1016/j.cmpb.2019.105020.

P75
The Evaluation Of The Morphological And Functional State Of Blood Lymphocytes In Patients With Abdominal Tuberculosis In Blood Smears
Viorika Akimova1, Nataliia Lapovets2, Lyubov Lapovets1
1 Danylo Halytsky Lviv National Medical Univercity, Lviv, Ukrenia, 2Research Institute of Epidemiology and Hygiene, Danylo Halytsky Lviv National Medical University, Lviv, , Ukrenia

Introduction: INTRODUCTION: Microscopic examination of blood smear remains an important diagnostic tool even in the age of molecular analysis and sophisticated laboratory tests. It is known that the blood lymphocytes' total count and their morphology characterized the immune reactivity. The nuclear/cytoplasm ratio reflects the level of cell metabolism and its functional state. Therefore, estimation of the size of lymphocytes can be used to define their functional status and immune reactivity of the patients’ bodies. The purpose of the work was to study the features of peripheral blood lymphocytes morphological structure in patients with abdominal tuberculosis. 

Methods: METHOD: Blood samples were collected from 65 patients with histologically confirmed abdominal tuberculosis (AT) (who entered the clinic during acute abdominal pain) before surgery and 30 healthy donors. It was studied structural features of lymphocytes, by percentage determination of large size (LL, d>10 µm), medium size (ML, d 7-9 µm), small size (SL, d< 6 µm) lymphocytes, large granule-containing lymphocytes (LGL) in blood smears stained by Romanowsky-Giemsa method using for examination light microscopy.
Results: RESULTS: Compared with healthy people the patients with AT has no difference in total leukocyte (6,85±0,6×109/l) and lymphocyte count (1,75±0,1×109/l). However, differences in the ratio of different morphological forms of lymphocytes were revealed. In AT is decreased the pool of the LL (25.5±0,5 % vs 32.3±0,6 %) and SL (40.3±0,6% vs 50.8±0,8%), while the pool of ML is increased (30.8±0,5% vs 14.1±0,1%). Also, higher content of BGL was determined in AT (4.4±0,1% vs 3.2±0,1%). The obtained results indicate an increase in the content of cytotoxic LGL and lymphocytes (ML) with high potential for immunogenesis, while the content of early postproliferative, mature, activated lymphocytes (SL) capable for recirculation and long-time circulating, aging lymphocytes (LL) is decreased. These results indicate activation of immunocompetent cells in AT, cytotoxic potential, that is confirmed by immunogrames in our previous investigation
Conclusions: CONCLUSION: The functional states of circulating lymphocytes can be previously estimated using differential morphological analysis of lymphocytes in the blood smear. The definition of lymphogram indicators does not involve the use of expensive equipment or special dyes and therefore can be widely introduced into practical medicine. The only necessary condition is the sufficient qualification of laboratory specialists

P76
Concurrent Beta Thalassemia Trait And Myelodysplastic Syndrome With Ringed Sideroblasts And Sf3B1 Mutation: Rare Case
YewSie Ang, BeeTang Yeo, GeeFung How, SimLeng Tien
Singapore General Hospital, Singapore, Singapore

Introduction: Beta thalassaemia is an inherited form of anaemia resulting from defective synthesis of beta globin chain, and the degree of anaemia is generally mild. Myelodysplastic Syndrome with ringed sideroblasts (MDS-RS) is an acquired clonal defect in the haematopoiesis characterised by anaemia, erythroid-lineage dysplasia, with or without multilineage dysplasia, < 5% blasts in bone marrow and (a) ≥5% ringed sideroblasts with SF3B1 mutation or (b) ≥15% ringed sideroblasts without SF3B1 mutation. 
Methods: Bone marrow smear examination using maygrundwald giemsa stain for morphologic classification of haematopoiestic cells. Prussian Blue iron stain was used to demonstrate the presence of haemosiderin (storage iron) in the bone marrow and stainable iron in erythroblasts (sideoblasts). Bone marrow examination was performed together with trephine biopsy, flow cytometry study, karyotyping and Fluorescence in-situ hybridisation (FISH for MDS). Molecular study was done to screened for SF3B1 mutations.
Results: Bone marrow examination revealed erythroid hyperplasia with < 5% blast cell, dyserythropoiesis, dysmegakaryopoiesis, hypogranular neutrophils and 48% ringed sideroblasts. Trephine biopsy revealed a hypercellular marrow with erythroid hyperplasia. Features consistent with the given background of thalassaemia and MDS cannot be excluded. 
 Flow cytometry showed erythroid hyperplasia with no significant trilineage dysplasia. Findings from  FISH were not supportive of MDS and there was no abnormal karyotype detected.  In view of the ringed sideroblasts, DNA sample extracted from  a bone marrow specimen was subsequently  screened for SF3B1 hotspot mutations and a nucleotide change of c.1997A>C (with a predicted K666T protein variant) was detected.
Conclusions: This is a first case reported in our institution of a patient with Beta thalassemia trait who later developed MDS-RS with SF3B1 mutation. This combination is uncommon with only limited cases reported in the literature. This highlights the importance of looking out for ring sideroblasts, followed by molecular study in the bone marrow evaluation of deteriorating anaemia in patients with beta thalassaemia trait. 

P77
Acute Myeloid Leukemia With T(8;16)(P11;P13) Translocation: A Typical Blast Cells Morphology
Marion Bareille
Université Catholique de Louvain, Hematology laboratory, Cliniques Universitaires Saint-Luc, Brussels, Belgium

Introduction: We report the case of a 12-year-old girl who presented to the hospital with fever and disabling arthralgia. Biological and clinical investigations revealed an acute myeloid leukemia (AML) with t(8;16)(p11;p13) translocation. The cytomorphological and immunophenotypic blast cells characteristics challenged us and prompted us to make a brief review of the literature.
Methods: Blood samples were drawn upon admission for a complete blood count, a biochemical profile and a hemostasis evaluation. A bone marrow aspiration was performed a few hours later for a cytomorphological examination, an immunophenotypic profiling, and a cytogenetic and molecular analysis. We then performed a systematic search in PubMed database concerning clinical, pathological and molecular features of AML with t(8;16)(p11;p13) translocation.
Results: Blood sample analysis upon admission revealed an inflammatory syndrome with an elevated C-reactive protein and an hyperferritinemia, as well as an early stage of tumor lysis syndrome with increased LDH and uric acid level. Blood count showed a grade 1 anemia, a grade 2 thrombocytopenia, and a white blood cells count within the reference range but with the presence of blast cells on the blood smear. Hemostasis evaluation pointed out elevated D-dimer levels consistent with disseminated intravascular coagulation (DIC). Cytomorphological examinations of the bone marrow aspirate smear showed a massive infiltration by a large blastic population showing abundant, vacuolated and finely granular basophilic cytoplasm as well as a regular nucleus with a spiked chromatin and large bluish nucleoli. Blast cells were also characterized by a positive myeloperoxydase (MPO) cytochemical staining. Last but not least, they exhibited the particularity of phagocyting platelets as well as elements of the erythrocytic lineage (Figure 1). Immunophenotyping by flow cytometry highlighted an abnormally high position of the blast cells in terms of cell complexity (SSC) on the biparametric graph SSC/CD45, with the expression of myeloid and monocyte orientation markers compatible with a FAB AML-M5 (CD34-, CD117-, HLADR+/-, MPO+, CD33+, CD4+, CD14-, CD11b-, CD15+, CD64+). Conventional karyotype revealed a t(8;16)(p11;p13) translocation. Characteristics retrieved in the literature concerning clinical, pathological and molecular features of this entity are summarized in Figure 2. 
Conclusions: Characteristics retrieved in the literature support our patient’s diagnosis of AML with t(8;16)(p11;p13) translocation. Onset and prognosis of this entity differ between adults and children, with a poor prognosis in adults and an intermediate one in children. 

Best Student Presentation Award Finalist
P78
Generation Of Artificial Images Of Leukocytes By Means Of New Generative Adversarial Networks (Leukocytesgan)
Kevin Barrera1, Anna Merino2, Andrea Acevedo1,2, Laura Boldú2, Angel Molina2, María Rodríguez-García2, Javier Laguna2, Jose Rodellar1
1Technical University of Catalonia, Barcelona, Spain, 2CORE Laboratory, Barcelona, , Spain

Introduction: Machine learning algorithms for automatic classification of blood cell images rely on well annotated datasets for their training. They have to be large in number and wide in covering a variety of cell classes. A recent methodology, generative adversarial networks (GAN), is increasingly used to generate synthetic images in several fields. Our goal is to automatically create images of leukocytes of peripheral blood starting from random numerical values.
Methods: We conformed a dataset of 4929 images of normal lymphocytes, 653 monocytes, 218 eosinophils, 236 neutrophils and 391 basophils. GANs learn the morphology of leukocyte images and create new images from random noise. We designed the system in two steps (Figure 1A). First, inserting random values in GAN A, a low-resolution fake image (90 x 90 pixels) is obtained. Second, this image is processed by GAN B to produce a high-resolution fake cell image (256 x 256 pixels) of one of the groups of interest. Morphology of artificial images was verified by expert clinical pathologists. In addition, they were automatically classified by a convolutional neural network (CNN) previously trained by our group with original leukocyte cell images.
Results: We generated 500 images of each cell class.  Subgroups of 100 images were delivered to five clinical pathologists of the Hospital Clinic of Barcelona to classify them according to their morphology. Results are shown in Figure 1B, highlighting 99.92% overall accuracy. All 2500 fake images were classified by the CNN. Figure 1C presents the corresponding confusion matrix, showing an overall accuracy of 95.72%.
Conclusions: Applying this innovative system, artificial images can be created with reasonable morphological fidelity, as verified with comparable accuracy by expert clinical pathologists and a trained classification algorithm. The main methodological feature is that, with the GAN models, the system learns the morphology of each cell group by transforming an initial set of random values into a realistic image with three color components. This allows to generate large databases in short time. In addition, the automatic verification of a batch of 2500 images takes a few minutes. A potential interest is to increase the number of cell images to facilitate the training of deep learning models.

P79
Chronic Myelomonocytic Leukemia: Testing A New Diagnostic Tool Using Sysmex Xn Analyzer Parameters.
Hanaa Bencharef
Chu ibn rochd casablanca , Casablanca, ON, Morocco

Introduction: Chronic myelomonocytic leukemia (CMML) is a clonal hemopathy characterized by ‘proliferative’ and 'dysplastic' variant. Its diagnosis remains difficult. The objective of this study is to test and validate a score to differentiate CMML from reactive monocytoses and thus reduce the number of blood smears in case of monocytosis.
Methods: A prospective and analytical study over a period of 3 months from December 21, 2020, in our laboratory . Were included patients> 18 years with a Monocytosis> 1G / L and> 10% of the leukocyte count.The study was carried out on a SYSMEX XN 9100 analyzer, . the blood smear: smearing, staining with MGG and microscopic reading were performed manually with the objective of confirming monocytosis and looking for signs of monocytic dysplasia. The data were collected and analyzed by the EXCEL software the MS was calculate using : Neutrophils to Monocytes ratio (Ne / Mo), absolute value of monocytes (Mo) and the Ne-WX parameter, according to this equation : : 1/(1 + exponential (-(11,623 + 0,026 * Ne-WX -1,385 * Ne/Mo + 2,714 * Mo))).
Results: 263 patients were collected with a sex ratio of 1.4. the mean value of white blood cells was 11 216 elements / mm3 [3 050-32 100], 1 492 elements / mm3 [1 000-5 020] for monocytes and 7 877 elements / mm3 [1 280-65 700] for neutrophilesas Table I.The MS was < 0.160 in 242 patients, and> 0.160 in 21 patients of which 9 patients showed signs of monocytic dysplasia on blood smear. In our study, the MS showed 100% sensitivity and 96.41% specificity with a negative predictive value at 100% Table II.
Conclusions: A Monodysplasia score was defined from the parameters of the SYSMEX XN analyzer, to allow the differentiation between a reactive monocytosis and a LMMC. Theefficiency of MS has already been confirmed in a validation cohort, which showed high sensitivity and specificity, as is the case in our study. The diagnosis of LMMC remains difficult, so the use of Ms will improve its detection and will decrease the number of unnecessary smears.

Best Student Presentation Award Finalist
P80
Expression Of Cd47 In Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma (Cll/Sll) By Immunohistochemical Stain
Brittany B. Coffman, Qian-Yun Zhang
University of New Mexico Department of Pathology, Albuquerque, NM, United States

Introduction: New advances in the field of immune checkpoint modulation include such agents as anti-CD47 monoclonal antibody.  CD47 (also known as integrin-associated protein) is a transmembrane receptor ubiquitously expressed in normal cells.  Binding of CD47 to its receptor located on macrophages (called SIRP-alpha or signal regulatory protein alpha) prevents phagocytosis of the cell by the macrophage.  This process serves as an immune check-point and allows for removal of potentially harmful pathogens and even cancer cells.   It has been identified that in the process of carcinogenesis, tumor cells up-regulate expression of CD47 in order to elude phagocytosis from the immune system, conferring survival benefit.  Identification of CD47 on tumor cells may provide prognostic information and as possible treatment, as anti-CD47 monoclonal antibodies are in clinical trials.  In this study we evaluated the staining patterns of CD47 in CLL/SLL.         
Methods:  4-μm-thick paraffin-embedded tissue sections from previously constructed tissue microarrays of CLL/SLL were baked and incubated with CD47 monoclonal antibody at a dilution of 1:100 (B6H12, Santa Cruz Biotechnology, Inc.) following antigen retrieval on a BenchMark Ultra instrument.  Slides were then reviewed by two separate hematopathologists to determine staining presence/absence, intensity (weak, moderate, strong), percent of cells staining, and staining pattern (nuclear, cytoplasmic, or membranous).  Cases of pathologist discrepancy were reviewed and consensus was reached.  This information was recorded in a REDCAp database and descriptive statistics were performed.
Results:  A total of 4 tissue microarrays containing 98 cases and 196 tissue sections were stained.  Of these 14 sections had insufficient tissue for review and were excluded.   Among CLL, 29% of cases were negative for CD47.  Most cases (42%) demonstrated weak staining intensity, 24% moderate intensity, and 9% with strong intensity.  The lymphoma sections demonstrated variable staining percentage, with 7% of cases demonstrating positivity in less than 25% of cells, 13% of cases in 25-50% of cases, and 25% in 75% or greater of cells. All cases demonstrated cytoplasmic staining pattern.
Conclusions: CLL demonstrates variable CD47 positivity and intensity, but uniform cytoplasmic staining pattern in all cases.  Our initial review indicates that assessment of CD47 expression in CLL may have future clinical implications. Additional assessment of CD47 expression patterns in other lymphomas are in process. 

Best Student Presentation Award Finalist
P81
Morphological Cues And Comparison Of Haematological Profile Between Neuroblastoma Patients With And Without Bone Marrow Infiltration - Study From A Tertiary Care Centre
Durga devi S1, Madhu Kumar Vengala1, Nabhajit Mallik1, Sreejesh S1, Narender Kumar1, Shano Naseem1, Prashant Sharma1, Man Updesh Singh Sachdeva1, Jasmina Ahluwalia1, Reena Das1, Amita Trehan2, Neelam Varma1
1Department of Hematology, PGIMER, Chandigarh, India, 2Department of Paediatrics, PGIMER, Chandigarh, India

Introduction: The bone marrow involvement (BM) in neuroblastoma represents the advanced disease and hence proper identification of BM infiltration is crucial for staging, prognosis and treatment purposes. The current study aims to evaluate the morphological features among BM involved cases and haematological profile between neuroblastoma patients with and without BM infiltration.
Methods: In this study, we included 115 cases of newly diagnosed neuroblastoma over the last five year and two months (January 2016 to February 2021). The clinico-pathological parameters were assessed. Statistical analysis was done using IBM-SPSS and P value < 0.05 was taken as significant.
Results: Of the 115 neuroblastoma cases, 54 (46.9%) showed BM infiltration.  The median age was 2years (range: 2months -12 years) and male to female ratio was 1.1:1.The median Hb, TLC and platelet counts were 95g/L (range: 54-143g/L), 9.6x109/L (range: 2.7-50.50x109/L) and 368x109/L (range: 24-1013.30x109/L) respectively. The most common clinical presentation was combination of abdominal mass, pain, fever, pallor followed by orbital swelling. Anemia was seen in 67(58.26%) of total cases, 40(74%) and 27(23.5%) of involved and uninvolved cases respectively. Leucopenia was seen in 3(2.6%) of total cases, 2(3%) and 1(1.6%) of involved and uninvolved cases respectively. Thrombocytopenia was noted in 16(13.9%) of total cases, 12(22.2%) and 4(6.5%) of involved and uninvolved cases respectively. Anemia and thrombocytopenia were significantly associated with bone marrow involvement. Infiltration was noted in both aspirate smears and trephine biopsy in 48 out of 53 cases, in trephine biopsy alone in 3 cases and in aspirate smear alone (absent in biopsy) in 2 cases. Infiltration pattern was diffuse in 30 (58.8%) cases and focal in 16(31%) cases. Rosetting, fibriilary neuropil like material and fibrosis was observed in 27(50%), 29(53.7%) and 26(48.1%) of involved cases. In addition necrosis and crushing artefact was observed in 9(16.6%) and 14(25.92%) cases respectively.
Conclusions: Among the hematological parameters, anemia and thrombocytopenia were found to be significantly associated with bone marrow involvement. Additional notable findings included presence of hepatosplenomegaly only in cases with and Opsoclonus myoclonus ataxia syndrome(OMAS) only in cases without BM infiltration.

Best Student Presentation Award Finalist
P82
Semi-Quantitative Scoring Of Myeloperoxidase-Stain Positivity Using Deep Learning
Haruhi Ida1, Tsukasa Sakashita2, Taiko Yamada2, Karin Masami2, Iori Nakamura1, Hiromi Masauzi3, Kazunori Okada3, Sanae Kaga3, Keiko Miwa3, Nobuo Masauzi3
1Graduate School of Health Sciences, Hokkaido University, Sapporo, Japan, 2Department of Health Sciences, School of Medicine, Hokkaido University, Sapporo, Japan, 3Faculty of Health Sciences, Hokkaido University, Sapporo, Japan

Introduction: Myeloperoxidase (MPO)-staining positivity is essential for the diagnosis of acute leukemia, but is still visually  evaluated.  We present a semi-quantitative scoring of MPO-staining positivity using texture description index (TDI) calculated from texture features that correlate with the density and distribution of MPO-positive granules.
Methods: MPO-stained leukocyte images from peripheral blood smears of 26 healthy volunteers (13 males and 13 females) were obtained using a digital camera.  The MPO-staining positivity of each image was scored as the integer (0–4) closest to the average score of the eight examiners.  The number of collected images was 1,510, 260, 3,159, 227, and 180, totaling 5,336, in order for Score 0 to 4, respectively.  Since the images for Scores 1, 3, and 4 were insufficient, these were increased 12-fold by rotating, horizontally, and vertically flipping using an in-house program.  Two groups (1,000 and 100) of images (totaling 5,000 and 500) were randomly selected for each score.   Six types of gray level co-occurrence matrices (GLCM) were generated for the 5,500 images, and nine TDIs were calculated from each GLCM.  The former 5,000 images and the latter 500 images were labeled with the nine TDIs and the correct MPO-score to be utilized as teacher and test data, respectively.  We repeated 100 training sessions of an original deep learning (DL) architecture and evaluated the classification performances.
Results: The accuracy of all test data, mean positive predictive rate, and sensitivity were 0.68, 0.5564, and 0.6333, respectively.  The MPO-scoring of eight examiners indicated a significant (χ2 = 9081.9, p< 0.0001) dispersion.  Therefore, we trained the same architecture using the teacher data labeled with only one examiner.  The performance indices improved to 0.742, 0.7523, and 0.742, respectively.
Conclusions: The observed results suggested that TDIs calculated from GLCM were useful for evaluating the possibility of MPO-stained cell images.  However, it is difficult to ensure the reproducibility and uniformity of MPO-score labels in teacher data.  Another automatic method to build the correct MPO-score label needs to be developed.

Best Student Presentation Award Finalist
P83
Standardization Of Discrimination Parameters By Artificial Intelligence-Based Segmentation Of Blood Cell Images
Hajimu Kawakami, Hirokazu Kurata
Sysmex Corporation, Kobe, Japan

Introduction: The judgment of blood cell morphology has been performed mainly by visual inspection, and based on the experience and ability of examiners. Recently, machine learning, including deep learning (DL), have been enabling the automated and objective blood image analysis, but the capability is still greatly influenced by the quality of teacher interpretation data. This is because the strict application of judgment criteria requires not only the digitization of parameters but also the accurate image recognition and measurement. In this study, we investigated a method for accurately recognizing and measuring cell components from blood cell images.
Methods: About fourthousand images of peripheral blood (PB) smears from normal volunteers, were obtained by DI-60 (Sysmex). (1) By combining Instance Segmentation (Mask R-CNN) and Semantic Segmentation (U-NET) methods, we created an algorithm to accurately identify areas of various cellular components, such as nuclei and cytoplasm. (2) The accuracy of area recognition was confirmed by F-score and IoU (​​intersection over ​​union) of the areas recognized by the algorithm and visual inspection. (3) The areas of each cellular components were measured, and the nuclear-cytoplasmic (N/C) ratio and the virtual cell and nuclear diameters were calculated. (4) The discrimination model was improved by 'Random Forest' method, using twenty-five discrimination parameters, and combined with deep learning (Ensemble method).
Results: (I) The comparison of the regions identified by the algorithm and the visual judgment indicated a good F value and IoU averages, as below. Lymphocytes/Monocytes F value: 0.979 (variance: 0.015), IoU mean: 0.960 (variance: 0.027) Neutrophils F value: 0.967 (variance: 0.019), IoU mean: 0.935 (variance: 0.020) (II) For thediscrimination of eight PBcell types, the accuracy of 0.98was achieved by using the Ensemble method, starting from the accuracy of 0.78 ofthe initial Random Forest model. Moreover, the analysis of judgment contribution rate revealed the five most significant parameters, as nucleus diameter, cell diameter, cytoplasmic blue color tone, and N/C ratio.
Conclusions: A large amount of accurate teacher data is required for the establishment of high-quality machine learning for the blood cell image analysis. In this study, using the newly developed segmentation algorithm and Ensemble method, we realized the objective measurement of cellular images with good reproducibility, as well as the standardization of cell type judgment.

P84
Detection Of Malignant Cells In Peripheral Blood Using Cellavision®Dm96
Javier Laguna1, Natalia Rakislova2, María Rodríguez-García1, Angel Molina1, Maite Rodrigo2, Miriam Cuatrecasas2, Anna Merino1
1CORE Laboratory, Biochemistry and Molecular Genetics Department, CDB, Hospital Clínic, Barcelona, Spain, 2Pathology Deparment, CDB, Hospital Clínic, Barcelona, Spain

Introduction: The term carcinocythaemia was proposed in 1976 for the finding of carcinoma cells in peripheral blood (PB). Since then, 49 cases have been reported in advanced cancer patients. We present carcinocythaemia finding in two patients with breast andgastric carcinoma. Besides, we report a third case of myeloid neoplasm. In this case, malignant cells were found in PB, being negative the bone marrow (BM) examination.
Methods: Case 1: A 65-year-old woman was diagnosed with breast carcinoma and bone metastases. Cell blood count (CBC) showed anaemia (haemoglobin 67 g/L) and thrombocytopenia (42x109/L). Case 2: A 32-year-old man was admitted for constipation, abdominal bloating and vomiting. He lost 20 kg in last months.CBC showed anaemia (haemoglobin 68 g/L) and thrombocytopenia (39x109/L). Case 3: A 75-year-old man was admitted with dyspnoea and cough. CBC showed anaemia (haemoglobin 72 g/L) and thrombocytopenia (52x109/L). CBC were obtained in ADVIA®2120i analyser (Figure 1) and PB was analysed using May-Grünwald-Giemsa staining. Cell images were acquired in CellaVision®DM96.
Results: Abnormal cells were detected in PB in all cases (Figure 2): Case 1: Presence of large epithelial-like cells with high nucleus/cytoplasm ratio and spongy nucleus with nucleolus. Immunocytochemical studies of CK19 and CKAE1/AE3 confirmed the carcinocythaemia. Case 2: Large epithelial-like cells were detected, showing prominent nucleoli, which were also present in BM. An abdominal computed tomography revealed a thickening of the gastric wall and bone metastasis. Gastric biopsy proved a diffuse gastric carcinoma. Immunocytochemical stains for Ber-EP4 and CKAE1/AE3 confirmed the carcinocythaemia. Case 3: Malignant haematological cells were observed on the edges of the smear, showing high nucleus/cytoplasm ratio and spongy nucleus with nucleolus. Some neutrophils contained green inclusions, indicating acute liver injury (AST: 150 U/L, ALT: 219 U/L and GGT: 872 U/L). BM showed absence of blast cells. Immunophenotype revealed 0.2% of CD34+ cells, and pathogenic ASXL1 mutation c.1762C>T was detected. Liver biopsy showed infiltration by haematopoietic elements (MPO+,CD33+,CD68+). Final diagnosis was myeloid neoplasia with hepatic infiltration. All patients died shortly after carcinocythaemia detection.
Conclusions: PB morphology can be used as potential minimally invasive diagnostic tool. Malignant cell detection in PB is a marker of dismal prognosis, useful for monitoring minimal residual disease.

P85
Analysis Of Severe Eosinophilia Cases Detected In The Laboratory During 2019
Javier Laguna, María Rodríguez-García, Angel Molina, Anna Merino
CORE Laboratory, Biochemistry and Molecular Genetics Department, CDB, Hospital Clínic, Barcelona, Spain

Introduction: Eosinophils are blood cells with an important role in parasite infection and mechanisms associated with allergy. Eosinophilia occurs when the eosinophil count in peripheral blood (PB) exceeds normal values (< 0.5x109/L). In this way, we can differentiate between mild (0.5-1.5x109/L), moderate (1.5-5x109/L) and severe eosinophilia (>5x109/L). In patients with eosinophilia, the first step is to determine if it is caused by a primary eosinophilia or is secondary to another process (reactive eosinophilia). Reactive eosinophilia is more frequent, highlighting allergy and parasite infections as common causes. However, there are other infrequent malignant diseases with eosinophilia, which should be considered. Our aim is to analyse the diagnoses of patients in which severe eosinophilia was detected in our clinical laboratory during 2019.
Methods: During 2019, we processed a total of 334,081 cell blood counts in the Advia®2120i analyser. High eosinophil counts were detected in 10,022 samples (3%), being mild eosinophilia the most common (9,302; 92.8%). Moderate (648; 6.5%) and severe eosinophilia (72; 0.7%) were less frequent. These 72 samples with severe eosinophilia (range: 5,3-21,1x109/L) were from 27 patients (11 patients with neoplasia and 16 patients with non-neoplastic causes) (Figure 1).
Results: In non-neoplastic group (16), the main cause of severe eosinophilia was parasite infection (eight patients, range: 5.7-13.2x109/L). In the remaining eight patients, the diagnoses were the following: three patients with Churg-Strauss syndrome (range: 5.3-10.3x109/L), three patients with idiopathic eosinophilia (range: 5.3-11.2x109/L), one patient with Kimura disease (6.2x109/L) and one patient with eosinophilic gastroenteritis (12.0x109/L). No patient with allergy and severe eosinophilia was found in this group. All patients with neoplasia (11) had remarkable leucocytosis, with values among 12.3-345.7x109/L. The following haematological malignancies were found in seven patients: 1) chronic myeloid leukaemia (CML) in four, showing eosinophil values between 6.1-21.1x109/L along with immature granulocytes and blast cells in PB; 2) B or T-cell lymphoma in three, reaching eosinophil values between 6.2-17.9x109/L. Finally, severe eosinophilia was found in four patients (range: 5.3-17.3x109/L) associated to pancreas, colon, lung and cervix malignant tumours.
Conclusions: Although parasite infection is one of the most frequent causes of eosinophilia, clinicians should take into consideration other diagnoses in cases of severe eosinophilia, including Churg-Strauss syndrome, Kimura disease, eosinophilic gastroenteritis and haematological or solid tumour malignancies.  

P86
An Evaluation Of The Operational Utilityof The Advanced Red Blood Cell Application Of The Sysmex Cellavision Di-60 Digital Morphology System
Richard McCafferty, Elaine Browne, Mairead Kearns
St James's Hospital, Dublin, , Ireland

Introduction: Manual microscopy is the gold standard method for examination of peripheral blood smears. However, it remains laborious, subject to subjectivity and lacking in accuracy of quantification, particularly in regard to red cell morphology. This study evaluates the operational utility of the Advanced Red Cell Application of the Sysmex Cellavision DI-60 digital morphology system, compared to manual reporting by experienced observers, using reporting criteria established by the International Commitee for Standardisation in Haematology.
Methods: 100 samples, 80 abnormal and 20 normal samples were analysed over the course of this study. Once a grading had been given by the Cellavision, a trained operator manually reclassified the cells where required to give a post-classification result. The gold standard method of light microscopy was compared to the preliminary grading given by the advanced red cell application and to the post-classification result after manual adjustment. Time trials were run and a questionnaire was completed by users to assess the operational utility of the application.
Results: Each abnormality yielded varied results, the most notable being for sickle cell samples and schistocytes. The DI-60 exhibited high sensitivity and specificity for sickle cells across all comparisons. However, while the analyser showed high sensitivity of 100% for schistocytes at the lower ICSH cut-off of >/= Grade 1, there was a lower specificity of 56% with a significant false positive percentage when the manual method was compared to the pre and post classifications. However, specificity for schistocytes improved to 92% with a lower sensitivity of 75% when the higher ICSH cutoff of >/= Grade 2 was programmed to the analyser. For other abnormal RBC morphologies such as target cells, tear-drop cells and acanthocytes, sensitivity was lower at 29%, 17% and 20% respectively, but specificity was very high at 98% or greater for all three.
Conclusions: The advanced red blood cell application of the DI-60 is a useful and user-friendly tool when used as a visual aid along with manual light microscopy. It brings greater objectivity and a potential for more standardised quantitation of RBC abnormalities, however it would benefit from further refinement. The high sensitivity for schistocytes may mean that there may be scope for this technology to be used to help further define the "true" lower cut-off percentage for schistocyte levels in normal subjects.

P87
Morphology And Automatic Recognition Of Atypical Lymphocytes In Peripheral Blood In Covid-19 Patients
Anna Merino1, Javier Laguna1, Laura Boldú1, Angel Molina1, Kevin Barrera2, Andrea Acevedo2, Jose Rodellar2
1CORE Laboratory, Hospital Clinic, Barcelona, Spain, 2Universitat Politècnica de Catalunya, Barcelona, Spain

Introduction: Peripheral blood (PB) smear review in COVID-19 patients shows the presence of atypical lymphocytes,hereinafter COVID-19 RL. This study aims to develop an automatic system for their objective recognition using convolutional neural networks (CNN).
Methods: 71 COVID-19 patients were included: 56 patients showing COVID-19 RL in PB (positive group), and 15 patients without (negative group). Mann Whitney U test was used for comparison of parameters. Samples were analysed with Advia®2120i and stained with MGG. Digital images were acquired with CellaVision®DM96. Two sequential CNNs were developed: one to distinguish between normal and reactive lymphocytes (RL); another to discern between COVID-19 RL and classical RL associated to other infections (Figure 1A). Sensitivity, specificity and precision were calculated. We assessed the system using lymphoid cell images from all COVID-19 patients and 30 patients with other viral infections.
Results: Median values (years) were 55±16 in the positive group, and 69±16 in the negative, p< 0.001. Haemoglobin and neutrophil/lymphocyte ratioshowed significant high values in the positive group (p< 0.002). Significant increased values ofneutrophils (p< 0.009), D-dimer (p=0.003) and procalcitonin (p=0.05) and decreased values of total protein (p< 0.001) and albumin (p=0.019) were found in the negative group. Patients requiring ICU admission were 27% in the negative group and 7% in the positive. Two patients of the negative group died and none in the positive. COVID-19 RL show larger size, deep blue cytoplasm and nucleus mostly in eccentric position with nucleolus (Figure 1B). Through two sequential CNNs, they were distinguished from normal and classical RL with sensitivity, specificity and overall accuracy of 90.5%, 99.4% and 98.7%, respectively (Figure 1C). In the proof of concept, all 56 smears (100%) were correctly classified as COVID-19 RL and all 15 smears (100%) were correctly classified as having normal lymphocytes. In the group with other viral infections, 28/30 (93.3%) were classified as smears having classical RL.
Conclusions: The automatic approach is useful for a more objective morphological recognition of COVID-19 RL, being able to classify smears in one of the groups under study. The presence of COVID-19 RL is consistent with abundant production of virus-specific T cells, thus explaining the better outcome of patients with these cells in PB.

Best Student Presentation Award Finalist
P88
Deep Learning Approach For Malaria Recognition Using Peripheral Blood Images Acquired With Cellavision Dm96
Angel Molina1, Jose Rodellar2, Laura Boldu1, Andrea Acevedo2, Edwin Santiago Alferez3, Anna Merino1
1Hospital Clinic of Barcelona, Barcelona, Spain, 2Technical University of Catalonia, Barcelona, Spain, 3Faculty of Natural Science and Mathematics.Universidad del Rosario , Bogotá, Columbia

Introduction: Several works have been done to assist in malaria diagnosis using machine learning techniques, but to date the presence of other inclusions that may interfere with malaria recognition has not been considered. In a previous work, we developed the first deep learning model using convolutional neural networks (CNN) able to differentiate malaria-infected red blood cells (RBC), not only from normal erythrocytes, but also from erythrocytes with other types of inclusions using images obtained with a conventional microscope. In this work we use images acquired with CellaVision DM96 to develop a new CNN model for malaria detection considering different types of RBC inclusions.
Methods: A total of 150 peripheral blood (PB) smear digital images were acquired with CellaVision DM96.  From them, 311,894 individual RBC images were segmented using thresholding techniques and watershed transform. An example of images obtained is shown in Figure 1. Since different classes of the dataset were highly unbalanced, images were downsampled to a total of 6,075 RBCs. Those images were used to train a ResNet34 architecture using transfer learning. The model performance was assessed using an independent test set of 74,304 images obtained from additional 31 smears.
Results: In the assessment, our CNN model obtained accuracy values of 99.6 % classifying malaria parasites and other RBC inclusions. Furthermore, sensitivity, specificity and positive predicted value for malaria recognition were 99.1, 99.97 and 95.15 %, respectively. The confusion matrix is shown in Figure 2.
Conclusions: The developed model represents a promising advance for the automation in the identification of patients infected with malaria.Unlike other automated applications, such as CellaVision Advanced RBC, which result in a high number of false positive cases due to misidentification, our model is very specific for the recognition of malaria. This work, using images obtained with CellaVision DM96, demonstrates the potential utility of our model in a real work environment.

Best Student Presentation Award Finalist
P89
Automatic Classification Of Bone Marrow Blood Cell Images Using A Two-Stage Deep Learning System With Automatic Segmentation And Semi-Supervised Learning.
Iori Nakamura1, Haruhi Ida1, Mayu Yabuta1, Shigeki Sato2, Shuichi Ota2, Naoki Kobayashi2, Koichi Matsumoto3, Keiko Miwa4, Nobuo Masauzi4
1Graduate School of Health Sciences, Hokkaido University, Sapporo, Japan, 2Sapporo Hokuyu Hospital, Sapporo, Japan, 3Cellavision Japan,inc., Yokohama, Japan, 4Faculty of Health Sciences, Hokkaido University, Sapporo, Japan

Introduction: Bone marrow (BM) cell differential counting is an important test, but it requires considerable labor and experience. To automate counting, we developed an automatic system for analyzing clustered BM cells using deep learning (DL).
Methods: We obtained cell images from 20 normal and pathologic anonymous BM blood films (May-Gruenwald-Giemsa-stained) using CellaVisionDM96 in digital slide mode. We first created a segmentation DL system by training with the original teacher data. Images totaling 35,656 were automatically segmented using the system. We then added correct labels to 20 images in each of the following 16 cell types of classification: proerythroblast/basophilic erythroblast, polychromatic erythroblast, orthochromatic erythroblast, blast, promyelocyte, myelocyte, metamyelocyte, band neutrophil, segmented neutrophil, eosinophile, basophile/mastocyte, monocyte, lymphocyte, plasma cell, mitotic-cell, and bare nucleus/artifact. These 320 labeled sets were used as the initial training data for the originally developed classification DL system.  We repeated 200 epochs of semi-supervised learning, which consisted of self-training and active learning, and we sequentially added labels to unlabeled images selected from the training data through semi-supervised learning by the system. New images were captured from five BM blood films that were not used in the training data. Fifty images of each of the 16 cell types were then selected to create the test data. The correct labels for these were determined by three qualified clinical laboratory technologists and a board-certified hematologist. The system was evaluated using the test data, and the accuracy, recall, and precision were calculated.
Results: After 18 rounds of semi-supervised learning, 18,775 images were labeled. The final classification accuracy for the entire set of test data was 0.964 (771/800), and the average recall and precision were 0.965 and 0.966, respectively.
Conclusions: We constructed a new segmentation system that can work even with hyperplastic BM blood films. This new classification system can classify 16 cell types at an accuracy of 0.964. In the future, we will increase the number of cases for training data and discriminate cell types. This should result in improved classification accuracy.

P90
Effects Of Recursive Additional Learning On Leukocyte Classification Ai Models By Deep Learning In Peripheral Blood Smear Screening
Hiroyuki Nozaka1, Honoka Harako2, Mae Miyazaki2, Suzuka Kaga2, Niina Sakaiya2, Kyouka Kudo2, Shou Kimura2, Manabu Nakano1, Miyuki Fujioka1, Kazufumi Yamagata1
1Hirosaki University Graduate school of Health Sciences, HIROSAKI, Japan, 2Hirosaki University School of Health Sciences, HIROSAKI, Japan

Introduction: Medical Artificial Intelligence (AI) is a next-generation medical technology that is able to provide a diagnosis based on EBM regardless of the experience of clinical laboratory technologists. The AI technology is characterized by learning large amounts of patient data diagnosed by experts based on years of experience. While many efforts have been started in medical image diagnosis, there are few research reports on techniques to improve accuracy of AI in hematological morphology screening. In this study, we examined recursive additional learning on leukocyte classification AI models with deep learning in peripheral blood smear screening, and clinical assessment was performed in peripheral blood smear screening.
Methods: The original image data set for training was consisted of 1331 typical mature leukocyte images. Data augmentation process was performed to increase the number of original images, and 3284 images were created (Data set A). ResNet-18 model was used for deep learning. The first CNN “Model A” was performed transfer learning with data set A. The subjects for evaluation were peripheral blood smear of 66 healthy persons (Test set A:51 cases, B:15 cases). The leukocytes misrecognized in the evaluation of “Test set A” were defined as “Data set B”. The second CNN was performed additional learning and finetuning with “Data set B”, and “Model B” was created. The evaluation was performed with both Test set A and B.
Results: The results of classification in Test set A were shown in Table 1, and classification in Test set B were shown in Table 2. It was shown by recursive additional learning with misclassified leukocytes that improvement in overall accuracy from 0.856 to 0.951 (+0.095) in test set A. Similarly, it was shown by recursive additional learning with misclassified leukocytes that improvement in overall accuracy from 0.736 to 0.889 (+0.153) in test set B.
Conclusions: It was cleared that recursive additional learning with misclassified leukocytes contributed to the improvement of leukocyte classification accuracy. Significant improvement was shown in segmental cells. It was suggested that recursive additional learning was greatly effective in improving the accuracy of AI model for leukocyte morphological analysis.

P91
Acute Myeloid Leukemia With Mutated Npm1 And Charcot-Leyden Crystals
Sophia Peng, Elona Turley
University of Alberta, Edmonton, AB, Canada

Introduction: Charcot–Leyden crystals (CLC) are most commonly found in eosinophilic inflammatory diseases such as asthma and parasitic infections, and in leukemic neoplasms associated with eosinophilia. There have been rare case reports of CLC without eosinophilia in acute myeloid leukemia (AML), but the significance of this morphologic finding is uncertain. Mutated NPM1 is the most common recurrent genetic lesion in adult AML and is prognostically important. In our practice, we noticed several cases of NPM1-mutated high-grade myeloid neoplasm or AML with CLC in the trephine biopsy. The objective of this case series is to describe the clinical, morphologic, and ancillary test findings of five cases.
Methods: Five cases were identified prospectively at diagnosis, and clinical information, morphology, and ancillary test results were reviewed.
Results: Patients ranged from 60-75 years-old. Four presented with bone pain. Peripheral blood findings included cytopenias (5/5), leukoerythroblastosis (4/5), and rare to occasional spherocytes (3/5). Circulating blasts ranged from 4-38%, with Auer rods in 1/5. Bone marrow aspirates were “dry taps” in all cases, necessitating repeat bone marrow sampling to establish the diagnosis in 3/5. Along with CLC, there was bone marrow fibrosis (MF-2/3 or greater) in 3/5 and necrosis in 2/5. Initial diagnoses were AML (2/5), high-grade myeloid neoplasm (2/5), and myeloid sarcoma (1/5).  Flow cytometry confirmed CD34-negative myeloid blast lineage in all cases. Ancillary testing are listed in table 1. One patient passed away prior to therapy. The remaining four underwent induction chemotherapy with subsequent disease relapse(s) (3/4) or undetermined status (1/4). Further study of bone marrow crystals confirmed autofluorescence and staining with Ziehl-Neelsen stain, which is characteristic of CLC (figure 1).
Case Diagnosis FLT3-ITD (allelic ratio) Karyotype Myeloid NGS
1 AML N normal N/A
2 myeloid neoplasm→AML Y (N/A) der(4)dup(4)(p16p15)del(4)(p16) N/A
3 MDS→AML Y (0.145) t(11;20)(q23;q11.2-13.1) DNMT3A, TET2
4 myeloid sarcoma N normal N/A
5 AML Y (0.598) normal DNMT3A
Table 1
Conclusions: Major findings in our cases of NPM1-mutated high-grade myeloid neoplasm or AML with CLC include bone pain at presentation, dry tap, and bone marrow fibrosis. In AML with a dry tap, trephine CLC could provide an important clue to disease subclassification.  A limitation of our study is the lack of NGS results for 3/5 cases. Future research underway includes a retrospective review of all recent AML with NPM1 mutation at our institution.

P92
Morphological Correlations Of Inv(16)(P13.1Q22) Or T(16;16)(P13.1;Q22); Cbfb-Myh11: Beyond French-American-British M4Eo - Single Center Retrospective Analysis Of The Last 5 Years
Ema Piloto1, Rita Ramalho1, Miguel Seruca3, Teresa Pereira2, Isabel Salazar2, Helena Alaiz2, Rui Barreira1
1Clinical Pathology Department, Instituto Português de Oncologia de Lisboa Francisco Gentil, EPE, Lisbon, Portugal, 2Hemato-Oncology Laboratory, Hematology Department, Instituto Português de Oncologia de Lisboa Francisco Gentil, EPE, Lisbon, Portugal, 3Clinical Pathology Department, University Hospital Center of Algarve, Faro, Portugal

Introduction: Although the most recent World Health Organization classification of tumours of haematopoietic and lymphoid tissues recognizes French-American-British (FAB) acute myelomonocitic leukaemia with abnormal eosinophils (M4Eo) as a synonym for acute myeloid leukaemia with inv(16)(p13.1q22) or t(16;16)(p13.1;q22); CBFB-MYH11, it also mentions that occasional cases with this genetic abnormality lack eosinophilia, have rare or difficult to identify abnormal eosinophils, and do not always present with morphologic myelo-monocytic differentiation. Furthermore, inv(16) and t(16;16) can also be present in therapy-related myeloid neoplasms.
Methods: We searched for all the de novo cases of acute leukaemia with inv(16)(p13.1q22) or t(16;16)(p13.1;q22); CBFB-MYH11, either primary neoplasms or therapy-related, diagnosed at our institution from January 2016 to December 2020. We then assessed the morphology of the patients’ bone marrow aspirate at the time of diagnosis.
Results: There were nine patients diagnosed with acute leukaemia who were positive for one of the above-mentioned genetic abnormalities (all of them were inversions). Four patients had a bone marrow aspirate morphology suggestive of acute myelomonocitic leukaemia with abnormal eosinophils (FAB M4Eo) and these were all primary neoplasms. Three other had a previous history of anti-neoplastic treatment and hence were diagnosed as therapy-related acute myeloid leukaemia. Finally, the remaining two patients both had bone marrow aspirates that revealed acute myeloid leukaemia without maturation (FAB M1) morphology. One of them was a 85 year old woman with a past history of anemia and thrombocytopenia, and the other one was a 55 year old man with no known history of hematological or oncological disease.
Conclusions: Among the six patients diagnosed with acute leukaemia not therapy-related, who were positive for inv(16)(p13.1q22), the bone marrow aspirate morphology FAB M4Eo could predict the diagnosis in 67% of cases (4 patients). The other two patients (33%) with acute leukaemia not related to previous anti-neoplastic treatment presented with morphological features of acute myeloid leukaemia without maturation (FAB M1).

P93
Cerebrospinal Fluid Microscopy Of A Patient With Second Relapse Of T(11;17)(Q23;Q21) Zbtb16(Plzf)-Rara Acute Promyelocytic Leukemia Eleven Years After Hematopoietic Stem Cell Transplantation
Ema Piloto1, Rita Ramalho1, Miguel Seruca2, Vanessa Lisboa1, Rui Barreira1
1Instituto Português de Oncologia de Lisboa Francisco Gentil, EPE, Lisbon, Portugal, 2University Hospital Center of Algarve, Faro, Portugal

Introduction: Acute promyelocytic leukemia (APL) is diagnosed in 8 persons per 10 000 000 population each year, accounting for 5-8% of acute myeloid leukemia cases. It is caracterized by the presence of the PML-RARA fusion gene; nevertheless, 1-2% of patients have variant gene translocations envolving retinoic acid receptor alpha (RARA) gene. The most frequent variant RARA translocation - t(11;17)(q23;q21) ZBTB16(PLZF)-RARA - has been associated with specific cytomorphologic characteristics.
Methods: We present the cytomorphologic characteristics of a t(11;17)(q23;q21) ZBTB16(PLZF)-RARA APL second relapse in the cerebrospinal fluid (CSF) of a patient, who had been treated with allogenic hematopoietic stem cell transplantation (HSCT) eleven years before.
Results: A 46 year old man followed in our hospital, who had been treated with allogenic HSCT in December 2008 for a first relapse of APL with t(11;17)(q23;q21) ZBTB16(PLZF)-RARA, presented with headache in January 2020. The investigation included a brain MRI, that showed several extra-axial lesions of dural basis, suggesting leptomeningeal involvement by leukemia. At that time, the blood cell count was unremarkable. CSF was colected by lumbar puncture and sent to our laboratory. The CSF was clear and the celular count was 50 nucleated cells/mm3. Differential count was performed after cytocentrifugation and May-Grunwald-Giemsa staining and showed 20% hypergranular promyelocytes, 6% myelocytes, 14% metamyelocytes, and 19% linfocytes. The promyelocytes had a single, non-lobed nucleus - some of them with one visible nucleolus - and hypergranulated cytoplasm, with some of the coarse granules obscuring the nucleus; no faggot cells were present.
Conclusions: The direct search for the original mutation of the patient’s leukemia by nested reverse transcriptase polymerase chain reaction (RT-PCR) in the CSF was positive for t(11;17)(q23;q21) ZBTB16(PLZF)-RARA. The bone marrow aspirate performed on the same day was non-representative for morphological analysis, however, nested RT-PCR confirmed the presence of bone marrow relapse.

P94
Assessment Of Morphological Anomalies Of Circulating Blood Cells In Covid-19
Dorra Rhim, Aya Chakroun, Nour Mziou, Sonia Mahjoub, Hela Baccouche, Neila Ben Romdhane
Biological hematology department of La Rabta hospital , Tunis, Tunisia

Introduction: severe cases of COVID-19 have a viral pneumonia and hyperinflammatory reaction which leads to acute respiratory distress and multi-organ failure. Hematological abnormalities in the complete blood cell (CBC) have been described in first studies of COVID-19 but blood smears are not performed systematically. Here, we aimed to describe morphological peripheral blood cells abnormalities in patients infected with SARS-CoV-2.  

Methods:  A prospective study enrolled COVID-19 in-patients of the Rabta hospital from November 2020 to February 2021. CBC was carried out on the sysmex XN-1000® hematology analyzer. Peripheral blood films were performed then colored by May-Grunwald-Giemsa stain. Examination of blood smears (BS) was realized by an experienced investigator. All data were analyzed on SPSS 19.0 software.
Results: There were 70 blood smears from in-patient with SAR-CoV-2 (from intensive care unit (n=38) and from other settings (n=32)). Results showed neutropenia (1.4%), hyperleukocytosis (64.3%), monocytosis (10%), lymphopenia (72.9%), and thrombopenia (28.6%). The most frequent morphological abnormality in circulating blood cells was hypergranular cytoplasm of neutrophil granulocytes (57,14%). Peripheral light blue agranular areas were noted in 22.8% of cases. Abnormalities of nuclear shape were observed in 21,4% cases. Assessment of blood smears also found neutrophils and monocytes containing multiple vacuoles (17%). Rare activated lymphocytes and large monocytes were found in some peripheral blood films. Platelet morphology also showed frequent anomalies, mainly consisting of giant platelets with different sizes (25%). No morphological abnormality was significantly associated to severe cases.
Conclusions: COVID-19 causes blood cells morphological abnormalities mostly related to inflammatory mechanisms and probably to perturbation of the myelopoiesis. To the best of our knowledge, blood smear examination has there was no predictive value of poor outcome in patients infected with SARS-CoV-2.

Best Student Presentation Award Finalist
P95
  Green Inclusions Detection In Peripheral Blood Neutrophils And Monocytes During The Period 2019-2020
María Rodríguez, Javier Laguna, Angel Molina, Francisca Bascón, Anna Merino
CORE Laboratory, Biochemistry and Molecular Genetics Department. Biomedical Diagnostic Centre. Hospital Clinic, Barcelona, Spain

Introduction:   Green neutrophilic inclusions (GI) have been reported in association with acute liver damage in different clinical conditions. These inclusions are increasingly been reported in the literature. We analysed clinical and biological findings in 16 patients in which we detected GI in our Hospital during the period 2019-2020

Methods: Blood smear (BS) review showed GI in sixteen patients (between 36 to 80 years-old), which were admitted in our hospital because different diseases (Figure 1). Full blood counts (FBC) were obtained using the Advia 2120i analyser, BS were stained with May Grünwald-Giemsa and images were acquired in the CellaVisionDM96.
Results: We detected the first two cases of GI in neutrophils in 2019 and we observed that the number of patients in which this abnormality was noticed during 2020 increased to 14. As it is shown in Figure 1, diagnoses of the patients were the following: infection or sepsis (11), liver transplant (2), ischaemic hepatitis (1), subarachnoid haemorrhage (1) and plasma cell leukaemia (1). Considering FBC in these 16 patients, low haemoglobin values was found in 12/16 cases, leukocyte counts were abnormal in 11/16 and platelet counts showed low values in 8/16. It is important to mention that AST and ALT were found increased in 14/16 patients, reaching values of 28,173 and 9,313 U/L, respectively in Case 15. Other relevant biological findings were: increased LDH in 15/16 patients, high lactate values in 11/16 and high CRP values in all cases in which it was determined. Increased prothrombin time was found in 11/16 patients. GI significance has been related to liver damage. Lipofuscin showed green, granular, cytoplasmic staining within hepatocytes with Giemsa stain and were morphologically similar to leucocyte inclusions, suggesting that they may be derived from lipid-rich lipofuscin released from necrotic hepatocytes.   In 12/16 patients (75%), clinicians were not aware of the acute liver failure when green inclusions were detected. In our group of patients, mortality rate was 56%.
Conclusions: Our data showed that laboratories should be proactive in screening for this GI and clinical pathologists should be aware of their connection with an acute liver injury. In our experience, green crystal detection in neutrophils predicted imminent death in most cases of patients with sepsis.

P96
Non-Promyelocytic Acute Myeloid Leukemia With Bilobed Blast Cells And Bundles Of Auer Rods &Ndash; A Case Report.
Ian Sarty1, Karen J. Harrison1, Rachelle Blackman2, Mary-Margaret Keating2, David M. Conrad1,3
1Department of Pathology, Dalhousie University, Halifax, NS, Canada, 2Department of Medicine, Dalhousie University, Halifax, NS, Canada, 3Department of Microbiology & Immunology, Dalhousie University, Halifax, NS, Canada

Introduction: Early recognition of acute promyelocytic leukemia (APL) is of paramount importance as it has a propensity to cause a coagulopathy that can be fatal to the patient. An APL diagnosis is definitively confirmed when t(15;17)(q24.1;q21.2) is identified in the leukemic cells. Fortunately, the characteristic morphologic features of APL make it relatively easy to detect by microscopy alone, which is much faster than a karyotype analysis is able to be performed. Specifically, nuclei are often kidney-shaped or bilobed, and bundles of Auer rods are virtually pathognomonic for APL.
Methods: We present a diagnostically challenging case of acute myeloid leukemia (AML) that mimicked APL by morphology.
Results: Bilobed blasts, as well as blasts with multiple Auer rods, were identified in peripheral blood. The 15;17 translocation was not detected by polymerase chain reaction; additional fluorescence in situ hybridization analysis ruled out the possibility of a complex or submicroscopic variant translocation involving chromosomes 15 and 17. The patient was ultimately treated with 3+7 and achieved morphologic remission.
Conclusions: This case underscores the potential diagnostic pitfalls in morphologic diagnosis of AML, and the importance of being cognizant of morphologic mimics of APL.

P97
Alinity Hq Plt Count Is Equivalent With The Cell-Dyn Sapphire Immunoplatelet Count In Severe Thrombocytopenia
Hussain Al-lawati1, Sumaiya Al-balushi 1, Safiya Al-Mahrazi 1, Alina Elena Pavel2, Muhammad Sehran3
1Royal Hospital, Muscat, Oman, 2Abbott, Wiesbaden, Germany, 3Abbott, Dubai, United Arab Emirates

Introduction: Platelet (PLT) counting technologies include impedance, optical, fluorescence and immunolabeling methods by automated hematology analyzers. Each technology is potentially differently impacted by interfering substances and conditions, therefore the accuracy of PLT enumeration is largely dependent on the underlying principle. Our goal was to compare PLT results obtained by the Alinity hq (Abbott) analyzer using multidimensional optical technology to those generated by CELL-DYN Sapphire (Abbott). Comparisons were performed with the CELL-DYN Sapphire Immunoplatelet method, which uses CD61 mAb and fluorescence flow cytometry to identify and enumerate PLTs, as well as with the CELL-DYN Sapphire optical PLT count.
Methods: One hundred sixty-nine patient samples with thrombocytopenia were selected from the routine workload of the laboratory of the Royal Hospital, Oman. Some samples had PLT morphological abnormalities, such as giant PLT, in addition to thrombocytopenia. Each sample was tested by CELL-DYN Sapphire and Alinity hq. A peripheral blood smear was made and reviewed for each sample. Quantitative results were compared with Passing-Bablok regression, Pearson correlation and Bland Altman statistics, using the Immunoplatelet method as reference. In addition, agreement for the PLT Clump flagging was determined by kappa statistics. 
Results: The median (min - max) PLT concentration was 49.4 (3.3-108.0) x 109/L with the Immunoplatelet method. Out of the 169 samples, the PLT count was invalidated by the analyzer in 19 samples with the Immunoplatelet method, and 42 and 44 samples with the Alinity hq and the CELL-DYN Sapphire PLTo method. Most of the invalidations was due to PLT Clump flags: 17 by Alinity hq and 22 by CELL-DYN Sapphire. The PLT Clump flagging agreement between the two analyzers was moderate, with kappa of 0.45 and total agreement of 88.8%. Alinity hq PLT counts showed strong correlation with both the Immunoplatelet and the CELL-DYN Sapphire PLTo methods. The bias between Alinity hq and Immunoplatelet results, as well as between Alinity hq and CELL-DYN Sapphire PLTo results was small and clinically insignificant. 
Conclusions: Alinity hq PLT counts, generated by the multi-dimensional optical technology were equivalent with the CELL-DYN Sapphire Immunoplatelet results in severely thrombocytopenic samples, implying that Alinity hq PLT counts are clinically reliable for making PLT transfusion decisions.  

P98
Ipf Analysis In Platelet Count Reduction On Hospital Admission Patients. Covid-19 Positive Versus Covid-19 Negative Cases
Giovanni Introcaso, Annalisa Cattaneo, Tiziana D'Errico, Lorenza Rondelli, Laura Cadau, Maria Luisa Biondi
Unit of Laboratory Medicine, Milan, Italy

Introduction: During the first outbreak of Sars Cov-2 we garnered hematological parameters of patients infected. Although there are many diagnostic tools, we claim the necessity to design a multimarker laboratory strategy for a timely characterization of Sars Cov-2. Aim of our study was to evaluate platelet parameters as potential available biomarkers. We focused on IPF, H-IPF and H-IPF/IPF ratio in patients admitted in emergency department (ED). 
Methods: We retrospectively analyzed 540 patients from February to April 2020. They were divided into two groups: (N=50) Covid-19 positive with molecular testing, (N=490) as reference population without molecular testing. Molecular diagnosis was made by Multiplex rRT-PCR from samples collected by nasopharyngeal swabs. Comparison of hematological parameters between groups was performed by a complete statistical analysis using SAS version 9.4. Continuous variables were compared using the t-test for independent samples. Variables not normally distributed are presented as median and interquartile ranges and were compared with the Wilcoxon rank sum test. Correlation study was conducted analyzing 9776 previous platelet counts before pandemic. Of Covid-19 positive patients, it was evaluated pre and post Covid-19 positivity PLT count. 
Results: We found abnormal platelet parameters as follows (Covid-19 group versus reference population): PLT(x109/L) 209(160-258) vs 236(193-279) (p=0.0239). IPF(%)  4.05 (2.5-5.9) vs 3.4 (2.2-4.9) (p=0.0576); H-IPF(%) 1.25 (0.8-2.2) vs 0.95 (0.6-1.5) (p=0.0171); (H-IPF/IPF Ratio)  0.32 (0.29-0.36) vs 0.29 (0.26-0.32) (p=0.0003) (median, IQR). It was observed a PLT reduction of nearly 50x109/L (median value) between pre Covid-19 and post Covid-19 sampling for each patient (N=43) (p=0.0194). Correlation study showed the follow results: Y= 0.4816X-0.6822, R2= 0.946 for the reference population before pandemic, data were extracted from general cardiological database (Figure 1); Y= 0.476X-0.5636, R2=0.947 for Covid-19 group (Figure 2).
Conclusions: We found a platelet count reduction in Covid-19 group confirmed by higher IPF and H-IPF values as turnover indices for platelet recovery. In particular, the new derived parameter H-IPF/IPF Ratio describes a higher number of PLTs with high fluorescence in Covid-19 group. It seems that correlation of IPF and H-IPF is similar into the two groups, suggesting that PLT reduction may trigger hematopoietic production even in Sars Cov-2 patients. These observations on platelet parameters shall be useful for a diagnostic application.

P99
Comparison Platelet Parameters In Sysmex Xn-2000 And Yumizen H2500
ANNA MARULL, JAVIER NIETO-MOARGAS, ORLANDO JIMENEZ, MERITXELL DEULOFEU, MAITE SERRANDO
LABORATORI CLINIC TERRITORIAL, GIRONA, Spain

Introduction: Platelets are irregular, disc-shaped elements in the blood responsible of blood clotting and the normal blood count range is 150×109/L to 400×109/L. Thrombocytopenia is a decrease in number of platelets in circulating blood; it can result from decreased or defective platelet production or from accelerated platelet destruction. Several automatedanalyzers are used in the clinical laboratories to perform complete blood cell analysis.  Sysmex XN-2000 provides three methods for platelets counting:  PLT-I which uses impedance, PLT-O based on optical methodology and PLT-F which is based on a Fluorocell fluorescent dye (oxazine).The Horiba Yumizen H2500 uses a Double Hydrodynamic Sequential System (DHSS) technology. Its significance is to ensure a linear flow of cells through electrical impedance method and optical measurement. The aim of this study is to compare the platelet count in Sysmex XN20 and Yumizen H2500 to know if there are significant differences.
Methods: A total of 649 samples were processed in January 2021 in Girona Territorial Laboratory. 101 of them were thrombocytopenias(plt < 100x109/L). Sample were run within 4 hours of collection and processed in Sysmex XN20 and Yumizen H2500 following the ISCH recommendations. Passing‐Bablok regression method and Bland-Altman analysis were performing to compare both methods. Spearman’s test was used to obtain the correlation coefficient.
Results: We compared impedance Yumizen H2500 platelets counts with Sysmex Impedance and Fluorescence (PLT-I and PLT-F). The results of Passing-Bablok, Bland- Almant analysis and Spearman rho correlation are shown in table 1. We observed a good correlation in all the parameters studied.  The Bland-Altman study of VPM presented consistently lower values in Horiba than in Sysmex. When we analyzed the bias in PDW Horiba presented constantly higher values.
Conclusions: New analytical equipment must be verified by the Clinical Laboratory to ensure that there are no major differences when changing methodology. Platelet counts using the Sysmex XN and Horiba Yumizen equipment have shown a correct correlation. Passing Bablok regression shows that the methods are not interchangeable. Bland Altman analysis shows that the differences may not be clinically significant.

P100
Alinity Hq Platelet Count Is Not Impacted By Severe Microcytosis
Fabrizio Papa1, Loredana Masi1, Sabatino Valente1, Zainab Mukhtar2, Gabriella Lakos2, Mark Lifson2
1Clinical Pathology Department, "San Giovanni Calibita" Fatebenefratelli Hospital, Rome, Italy, 2Abbott Laboratories, Santa Clara, CA, United States

Introduction: Impedance technology has been previously shown to overestimate platelet (PLT) count in samples containing low volume red blood cells, while the fluorescent PLT count by Sysmex has been suggested to be unaffected by the presence of microcytes. Our goal was to assess the accuracy of the Alinity hq (Abbott) multi-dimensional optical PLT counting technology in relation to the mean cell volume (MCV) of red blood cells and compare its performance to Sysmex XN impedance (PLT-I) and fluorescence (PLT-F) methods.
Methods: Patient samples were tested in a clinical study comparing the performance of Alinity hq and Sysmex XN-3000. The cohort (n=464) included anemias of various etiologies, such as iron deficiency, anemia of chronic disease, folate or B12-deficiency, thalassemia trait, blood loss, etc. The remaining cohort consisted of routine samples from the laboratory’s patient population. PLT results were compared across various MCV ranges.
Results: PLT concentrations ranged from 6.56 to 947 x 109/L, and MCV from 40.9 to 123.0 fL. Alinity hq PLT count demonstrated good overall correlation and agreement with the Sysmex PLT-I and PLT-F results, with stronger correlation compared to PLT-F (correlation coefficients of 0.98 and 1.00, respectively). Bland-Altman analysis showed an overall positive bias of 7.2 x 109/L between Alinity hq PLT and Sysmex PLT-F values. A mean bias of 9.3 x 109/L was observed by Sysmex PLT-I compared to Sysmex PLT-F results. Samples were categorized into three ranges: normocytic and macrocytic (MCV > 80 fL), microcytic (MCV 65-80 fL), and severely microcytic (MCV < 65 fL). Bland Altman analysis revealed an increasing mean bias of 6.0, 14.4 and 35.2 x 109/L between Sysmex PLT-I and PLT-F results in the three MCV groups, but a consistent mean bias of 7.0, 7.1 and 9.2 x 109/L between Alinity hq and PLT-F. Spearman correlation demonstrated a highly significant relationship between the Sysmex PLT-I vs PLT-F count difference and % microcytosis (p< 0.0001), implying that microcytosis is related to the overestimation of the PLT count by the Sysmex PLT-I method.
Conclusions: Alinity hq PLT counts are equivalent with Sysmex PLT-F results in severely microcytic samples, suggesting that presence of cells with low MCV does not impact PLT enumeration by the multi-dimensional optical technology.

P101
Abnormalities In Tissue Factor Pathway Inhibitor Levels In Commonly Encountered Platelet Function Disorders
S. Tasneem, T. Sharma, C.P.M. Hayward
McMaster University Medical Center, Hamilton, ON, Canada

Introduction: Tissue factor pathway inhibitor (TFPI) is a key regulator of the extrinsic pathway of coagulation that influences thrombin generation in vitro and in vivo. Megakaryocytes synthesize TFPI, which is released from platelets upon agonist stimulation. While plasma and platelets respectively contain 90% and 10% of the blood TFPI, unlike platelet TFPI, plasma TFPI is truncated and has much lower anticoagulant activity than platelet TFPI, which is mainly the full length protein. As animal studies indicate that platelet TFPI modulates bleeding and thrombosis, we undertook an evaluation of TFPI in platelet function disorders (PFD) to gain new insights on these disorders.
Methods: The study was conducted with approval of the Hamilton Integrated Research Ethics Board and written informed consent. Samples were collected from general population controls (n=15-23 tested/assay)and from patients (n=25-26/assay) that were part of our recently reported, well characterized PFD cohort (https://onlinelibrary.wiley.com/doi/full/10.1002/rth2.12374), with increased bleeding risks and increased bleeding scores. PFD cases evaluated had non-syndromic dense granule deficiency or impaired platelet aggregation responses due to RUNX1 haploinsufficiency or uncharacterized molecular causes. Participants were tested by calibrated automated thrombogram (Hemker method) to determine endpoints of TG. TFPI levels in plasma and platelets were evaluated by ELISA and by droplet digital polymerase chain reaction analyses of platelet RNA.
Data were analysed using Mann-Whitney tests and linear regression to look for relationships between findings, and between findings and bleeding scores.
Results: Although plasma TFPI levels [TS1] were not significantly different for PFD versus control participants, platelet TFPI levels [TS2] were significantly reduced in the PFD participants (respective findings as median [range] in ng/mg platelet protein: Controls 17.2 [8.7-25.9]; PFD 13.6 [7.8-24.2]). However, platelet TFPI transcri[TS3] pt levels of PFD participants were similar to controls. Neither plasma or platelet TFPI levels of PFD participants showed significant association to in vitro TG endpoints  (R2 ≤0.53, p≥0.07). However, among PFD participants, platelet TFPI levels showed significant, positive association to bleeding scores (R2=0.23; p=0.01).  [TS1]Tested in 25 PFD participants, and 19 controls  [TS2]Tested in 25 PFD participants and 15 controls  [TS3]Tested in 23 controls, and 26 PFD
Conclusions: Our study is the first report of platelet TFPI abnormalities in PFD. The association that we found between platelet TFPI levels and PFD bleeding, as quantified by bleeding scores, is quite interesting and suggests that platelet TFPI levels influence bleeding in commonly encountered PFD.

P102
Comparison Of Fluorescent Spot Test With Quantitative Enzyme Assay In Detecting Glucose-6-Phosphate Dehydrogenase Deficiency
Sarah Abdul Halim, Shafini Mohamed Yusoff, Rosnah Bahar, Wan Zaidah Abdullah
UNIVERSITI SAINS MALAYSIA, KOTA BHARU, Malaysia

Introduction: Glucose-6-Phosphate Dehydrogenase (G6PD) is an enzyme that catalyses the first reaction in the pentose phosphate pathway and is the only source of NADPH in red blood cells. G6PD deficiency is the commonest enzymopathy and is inherited via X-linked inheritance. Males are either hemizygous normal or hemizygous deficient, whereas females are either homozygous normal, homozygous deficient or heterozygous. In Malaysia, the neonatal screening program was introduced in 1980 and uses Fluorescence Spot Test (FST), a semiquantitative method. Although inexpensive, FST often fails to differentiate heterozygous with homozygous normal females.
Methods: This was a cross sectional study comparing FST and quantitative assay (BioSensor1 careSTARTTM ) done in Hospital USM. Both FST and the quantitative assay were performed on 455 cord blood samples taken at birth and sampled randomly during the study period. Parameters including gestational age and gender were collected from patients' records.
Results: 52.3% of our samples were female and 47.7% were male. The majority of our patients were Malay (97.4%) whilst the remaining were of Chinese, Arab, Thai and Rohingya ancestry. 93.4% of the samples were term neonates, 5.9% were moderate to late preterm (< 37 weeks) and 0.7% were very preterm (< 32 weeks).The mean cord blood G6PD level for neonates with normal FST was 7.0 U/gHb  ± 1.2 SD and  the mean cord blood Hb level was 16.5 g/dL ± 2.8 SD. The frequency of G6PD deficiency using FST and BioSensor1 in males was 7.4% and 13.8% respectively, whilst in females, the frequency was 2.5% with FST but increased to 19.7% using BioSensor1. The overall prevalence of G6PD deficiency was 4.8% using FST but significantly increased to 16.9% with BioSensor1. Using independent T-Test, there was no significant difference in mean G6PD level between term and preterm neonates (p value 0.138) however this could be due to small sample size.
Conclusions: Quantitative enzyme assay can detect more heterozygous females than FST, causing the discrepancy in overall prevalence of G6PD deficiency tested with different methods. It is prudent to test female neonates with quantitative enzyme assays to avoid misclassifying heterozygous females. Larger study with heterogenous population of premature neonates are needed to evaluate mean G6PD level of preterm neonates.

P103
Knowledge Of Sickle Cell Disease And Hemoglobin Types; A Rural Community Study In Nigeria
Patience Akpan, Ada Onete
University of Calabar, Calabar, Nigeria

Introduction: Sickle cell disease (SCD) is a major public health problem in Nigeria with carrier prevalence of 25%. The disease runs a chronic course; characterized by recurrent ill-health, progressive organ damage and shortened life-span. The study aim is to assess the level of knowledge and awareness about SCD and determine the hemoglobin types of indigenes of Ijom, Abayong, a rural community in Biase Local Government Area, Cross River State, Nigeria. 
Methods: A structured questionnaire was administered to two hundred (200) indigenes of Ijom Abayong to ascertain their level of knowledge/awareness towards SCD. Hemoglobin electrophoresis by Cellulose Acetate was used to determine their Hemoglobin (Hb) types. Data obtained were analyzed using Chi-square with significance set at P< 0.05.
Results: Results showed that fifty percent (50.0%) of the respondents were within the ages of 16-24 years with female comprising 55.5% and singles making up 52.5%. Educational level was 10%, 55% and 35% for tertiary, secondary and primary respectively. Sixty and a half percent of respondents had heard of sickle cell disease while 50.5% did not know the cause and 89.0% had not known anyone with the disease. Only 47.0% of respondents knew about hemoglobin type and 68.5% of those married did not check their Hb type before marriage. The Hb type of the study population comprised Hb AA (80.5%) and Hb AS (19.5%); no Hb SS was recorded. Among the indigenes of Ijom Abayong, 50.0% did not know how SCD was acquired with 42.5% affirming that it is through inheritance; those that did not know how it is diagnosed were 47.5%. Most respondents (62.0%) did not know how to prevent SCD while those who knew about it through school, internet and radio made up 33.5%, 18% and 17% respectively. Respondents who thought none of the children had a chance of inheriting SCD from a carrier parent were 26.0% while others thought the chance was ‘all’ (23.5%), ‘half’ (26.5%) and one fourth (12.5%) of the children; 11.5% did not know. Similar values were observed for the chance of inheritance from both parents with AS type. 
Conclusions: It is concluded that participants demonstrated a fair level of knowledge with a favorable absence of Hb SS.Public health education is recommended.

Best Student Presentation Award Finalist
P104
A Novel Optical Chip Aggregometer To The Preliminary Evaluation Of The Effects Of Combined Anesthetics On Erythrocytes
Analía Alet1, Martín Toderi1,2, Marcus Batista1, Bibiana Riquelme1,2
1Facultad de Ciencias Bioquímicas y Farmacéuticas (UNR), Rosario, Argentina, 2Física Biomédica, IFIR (CONICET-UNR), Rosario, Argentina

Introduction: Human red blood cells (RBC) aggregate face-to-face building linear structures that resemble a stack of coins (rouleaux). RBC aggregation can be analyzed by recording the intensity of transmitted light through a blood sample in real-time with an Optical Chip Aggregometer. It has been shown that drugs commonly used in anesthesia (Propofol (P), Remifentanil (R), and Vecuronium (V)) could alter the RBC aggregation phenomenon. In this work, we study the in vitro action of these drugs in combination to determine their effect on the RBC kinetics aggregation.
Methods: Fresh blood samples were collected by venipuncture in sterile vials containing EDTA as anticoagulant. Samples were incubated at 37ºC, diluting one volume of blood with one volume of physiologic solution added with the corresponding anesthetic drug (P: 4 mg/mL whole blood; V: 0.15 mg/mL plasma; R: 10 ng/mL plasma). Several syllectograms were obtained with the Optical Chip Aggregometer, and the aggregation kinetics curves were analyzed to obtain the normalized parameters.
Results: The parameters t1/2 (time required to reach 50% of the total amplitude of transmitted light), Amp1/2 (normalized light intensity at t1/2), AI (Aggregation Index, normalized area under of the syllectogram curve), and M-Index (integral of the syllectogram curve), were defined to assess them computationally (Table 1).     Table 1. Kinetics aggregation parameters (Mean±SD)
RBC Treatment t1/2  (s) Amp1/2  (a.u.) AI (a.u.) M-Index (a.u.)
C 28.5±0.4 88±1 0.82±0.01 32.0±0.5
P 34.0±0.8*** 84±1 0.80±0.02 31.6±0.7
R 35.0±0.8*** 90±1* 0.84±0.01 33.4±0.4*
V 28.5±0.5 89±2 0.82±0.01 32.9±0.5
PR 35.0±0.8*** 85±2 0.79±0.02 31.4±0.7
PV 30.5±0.5** 88±2 0.81±0.02 32.5±0.6
RV 24.5±0.3*** 90.9±0.9* 0.842±0.009* 33.7±0.3**
PRV 22.5±0.2**** 93.2±0.7** 0.856±0.007** 33.4±0.3*
“*” denotes p-value by Student’s t-test: *(p< 0.05), **(p< 0.01), ***(p< 0.001), ****(p< 0.0001) a.u. arbitrary unit     t1/2 variations showed that anesthetics had a synergistic behavior. The AI was significantly higher for RV and PRV, indicating that under these treatments’ RBCs present a higher amount of RBC aggregates.
Conclusions: In this preliminary study, it could be observed that anesthetics in combination can alter the aggregation parameters of RBC in vitro. The significant variations observed in the aggregation parameters indicate a synergistic effect of the anesthetics studied on the erythrocyte membrane.

P105
Delayed Hemolytic Transfusion Reaction Identified With Sickle Cell Crisis Recurrence: A Report Of Two Cases
Nino Balanchivadze, Christopher Willner, Khaled Almadhoun, Zaher Otrock
Henry Ford Hospital, Detroit, MI, United States

Introduction: Sickle cell disease (SCD) patients may develop delayed hemolytic transfusion reactions (DHTR), which can be life-threatening complications of red blood cell (RBC) transfusion. In this report, we describe two cases of DHTR identified during recurrent acute painful sickle cell crisis.
Methods: Report of two clinical cases
Results: Case 1: A 30-year-old African American (AA) male with history of hemoglobin SC disease presented to an outside hospital with chest pain, and pain in his hips and thighs bilaterally. Initial hemoglobin (Hgb) level was 8g/dL. His antibody screen was negative; 2 units of packed RBC were transfused. Patient was managed with intravenous fluids and opioids. He clinically improved and was discharged. He presented 14 days later to our hospital with the same pain symptoms and shortness of breath. Now his Hgb was 9g/dL; however, his anemia worsened to Hgb 5.7g/dL within 2 days. Antibody screen was positive for anti-C and anti-Fy(a) antibodies. Direct antiglobulin test (DAT) was negative. His lactate dehydrogenase (LDH) increased from 481 to 786 IU/L. He was transfused with 2 RBC units and was discharged 1 week from presentation. He was doing well 8 months after presentation with Hgb 12.2g/dL. Case 2: A 28-year-old AA female with HgSS SCD presented to the hospital with pain in the lower back, upper thighs and bilateral hips. Her vital signs revealed tachycardia, but her physical examination was otherwise unremarkable. Laboratory investigations were notable for Hgb 5.9g/dL and LDH 586 IU/L. Her antibody screen was negative. She was treated conservatively with hydration and pain control and was transfused with 2 RBC units. She was discharged home with Hgb 8.2g/dL. She presented back to the hospital after 4 days with ongoing pain in addition to shortness of breath and anemia, hemoglobin 5.4g/dL and rising LDH to 1881 IU/L. Her antibody screen was positive with anti-Jk(b) specificity. DAT was negative. She was managed with IV fluids, pain medications, and was transfused with 2 compatible RBC units with good response.
Conclusions: Our patients presented early after recent sickle cell crises management and after blood transfusions with evidence of hemolysis and drop in Hgb levels. DHTRs should be considered in patients presenting with early recurrence of sickle cell crisis and history of recent blood transfusions.

P106
Hemoglobinosis D-Punjab In A Moroccan Family
Hanaa Bencharef
Chu ibn rochd casablanca , Casablanca, ON, Morocco

Introduction: Hemoglobinosis D, also called Hemoglobinosis Los Angeles or Punjab, is a structural variant derived from a point mutation of the beta-globin (HBB) gene in the first base of codon 121,  with the substitution of glutamine for glutamic acid (Glu> Gln) in the beta globin chain. It is transmitted in autosomal recessive fashion and occurs in different forms including heterozygous hbD trait. It is widespread worldwide, with dominance in the Punjab region of the northwest of India, unlike Morocco where it is still very rare and poorly studied. We describe the first case of heterozygous HbD-Punjab in a Moroccan family, discovered following microcytic hypochromic anemia in a 6 years old child. This case highlights the possibility of the appearance of rare phenotypes within our multiethnic population, and emphasizes the importance of precise genetic studies to plan genetic counseling . 
Methods: It is about a child of Moroccan origin, aged 6 years, without concept of consanguinity in the parents, who was referred to us for the exploration of a microcytic hypochromic anemia with normal ferritinemia, the clinical signs were cutaneo-mucous pallor associated with chronic asthenia. Red cell indices at the hospital revealed the following: Hb 9 g/dL, RBC 2.98 X106/uL, MCV 73fL, and MCHC 30,9 g/dL. Blood smear showed marked anisopoikilocytosis with microcytic hypochromic red cells, without any sickle and boat shaped cells. The Hb electrophoresis done at pH 8.5 showed a variant band at hemoglobin S position, Further analysis on high performance liquid chromatogtaphy (HPLC) showed presence of Hb D variant at 39,1 % also checked by capillary electrophoresis. The parent’s blood samples were sent to our institute for   testing, objectifying a normal profile of hemoglobin electrophoresis in the mother and the presence of 40.7 % of HbD in the father of the child, therefore we were able to fully determine the pattern of inheritance.
Results:
Conclusions: The Hb D-Punjab is relatively widespread in the world, However, its remains extremely rare in Moroccan family and not documented in the literature. The diagnosis of this hemoglobinopathy, can be difficult, especially in asymptomatic cases, and it uses a variety of tests, including a phenotypic exploration with electrophoretic techniques and chromatography as well as a genetic study.

P107
Acute Hemolysis With Basophilic Stippling In Hospitalized Patient With Severe Covid-19 Infection
Mingyi Chen
UT Southwestern Medical Center, Dallas, TX, United States

Introduction: Severe acute respiratory syndrome coronavirus 2, coronavirus disease 2019 (COVID-19)-induced systematic infection with significant complications impacting on the hematopoietic system and hemostasis.  
Methods: We retrospectively analyzed hematological parameters of hospitalized COVID-19 patients with request for peripheral blood smear review from April to December of 2020.  The laboratory data in correlation with clinical outcomes was searched from electronic medical record. 
Results: A majority of 53 out of these 102 patients presented with severe dyspnea and high fever, and was admitted to intensive care unit (ICU) after being confirmed positive for SARS-CoV-2. Peripheral blood smears were also reviewed for morphological abnormality in correlation with laboratory data and clinical manifestations.  At admission, the majority of patients (72%) presented with lymphopenia and occasional plasmacytoid reactive lymphocytes.  Some of the ICU patients (14) developed severe hypoxic respiratory failure and was intubated.  After treatment with convalescent plasma, and hydroxychloroquine and azithromycin, the clinical course was often complicated with acute renal failure, disseminated intravascular coagulopathy (DIC).  The CBC data showed average hemoglobin 10.4 g/dl, hematocrit 36.5%, white cell count of 9.0x109/L (neutrophils 80.0% and lymphocytes 9.6%), platelets 66.0x109/L.   The typical features of peripheral blood smear analysis included normocytic anemia with increased polychromasia, circulating nucleated red blood cells, basophilic stippling, left shifted granulocytes with toxic granulations, lymphopenia and thrombocytopenia. 
Conclusions: Our study reveals that severe lymphopenia, relative neutrophilia with left shift granulocytes, hemolysis with nucleated red blood cells and raised LDH are associated with ICU admissions and severe acute respiratory distress syndrome (ARDS).  Acute hemolysis in COVID-19 patients is a common complication and adversely affects clinical outcome, especially in older patients with pre-existing comorbidities.

P108
Hba2 Yialousa And Β-Thalassemia&Ndash; A Case Report Of Compound Heterozygosity The Challenges Of External Quality Assessment Program Strials
Karanini Ferreira1, Ana Paula Azevedo1,2, Ana Batalha Reis1, Flora Meireles1, Cândido Silva1, Armandina Miranda3, Paula Faustino3, João Faro Viana1
1Clinical Pathology Department, HSFX, CHLO, Lisbon, Portugal, 2Centre for Toxicogenomics and Human Health (ToxOmics),Genetics,Oncology and Human Toxicology/NOVA Medical School/Faculty Medical Sciences,UNL, Lisbon, Portugal, 3Instituto Nacional de Saúde Dr. Ricardo Jorge (INSA), Lisbon, Portugal

Introduction: Hemoglobin A2 (HbA2) Yialousa is the most common δ-thal mutation found in the Mediterranean region. Although most variants are asymptomatic, it is very important to detect HbA2 variants in the premarital screenning diagnosis, because their interaction with β-thalassemia may result in a normal HbA2 value, leading to misdiagnosis of coexisting β-thalassemia. We report a clinical case of double heterozygosity for Yialousa HbA2 variant  and β-thalassemia, included in a trial of the Portuguese External Quality Assessment Program in Haemoglobinopathies  from Instituto Nacional de Saúde Dr. Ricardo Jorge (INSA), in which our laboratory participates. Our aim is to highlight the role in the specialized training of professionals, enabling access and knowledge for very rare cases in clinical practice, but so important to be diagnosed in order to prevent severe forms of hemoglobinopathies.
Methods: 35-year-old man, asymptomatic, no relevant clinical past history, referred to hematology consultation due to the presence of maintained slight anemia, microcytic hypochromic.
Results: Analytically: Hb-12.4g/dL; MGV-69.9fL; MGH-22.1pg, normal serum iron, HbA2 and HbF values, and normal α-gene study. DNA sequencing: β-globin gene mutation in heterozygoty and an additional mutation in δ-globin gene, led to diagnosis of β+ thalassemia coinherited with HbA2 Yialousa). His wife had β+thalassemia.
Conclusions: In β+thalassemia patients, co-inheritance of a δ-globin gene mutation, such as HbA2 Yialousa, although clinically asymptomatic, prevents the expression of biochemical phenotype characteristic of β+thalassemia, elevated HbA2 levels(≥3,5%), because of a thalassemia effect or instability, leading to normal/borderline HbA2 values. Stable mutations results in two HbA2 fractions of about half of the expected value. These variants make the diagnosis of βthalassemia more difficult to reach, only through molecular studies. Considering that this couple involves the association of one member with βthalassemia and the other with βthalassemia associated with a HbA2 Yialousa, there is an increased risk of severe hemoglobinopathy for the descendants (25% in each pregnancy). The trial in which this case was included, gave us the opportunity to look at and to study a very interesting and extremely rare clinical situation, very important in the screenning programmes  for preventing severe  forms of hemoglobinopathies.

P109
New Haematological Parameters Provided By Mindray Bc-6800Plus Analyser Useful In The Differential Diagnosis Of Anaemia
Javier Laguna, Laura Boldú, Angel Molina, María Rodríguez-García, Francisca Bascón, Claudia Carballo, Saray García, Anna Merino
CORE Laboratory, Biochemistry and Molecular Genetics Department, CDB, Hospital Clínic, Barcelona, Spain

Introduction: Anaemia is a common condition with varied causes: nutritional deficiencies, haemoglobinopathies and congenital or acquired haemolytic. New parameters involving red blood cells (RBC) in addition to the classical ones provided by new analysers might be useful to improve the differential diagnosis of anaemia. Our aim was to analyse the following new parameters provided by Mindray BC-6800Plus analyser and determine their utility in the differential diagnosis of anaemia: microcytic (MICRO), macrocytic (MACRO), hypochromic (HYPO) and hyperchromic (HYPER) RBC, mean reticulocyte hemoglobin content (MCHr), immature reticulocyte fraction (IRF), mean reticulocyte volume (MRV) and fragmented erythrocytes count (FEC).
Methods: In this study, samples from 30 healthy individuals and 42 patients with anaemia (11 iron-deficiency, 10 thalassemia, 8 post-haemorrhagic anaemia, 7 aplastic anaemia and 6 megaloblastic anaemia) were collected in EDTA and analysed on BC-6800Plus. Mann-Whitney U test was used considering statistically significant p-values< 0.05.
Results: In iron-deficiency and thalassemia, MICRO, HYPO, MCHr and FEC values were significantly higher (p< 0.001) than in healthy controls. We found significant higher values of IRF, MICRO and MRV in thalassemia patients than in iron-deficiency (p-values of 0.038, 0.011 and 0.016, respectively). In contrast, HIPO values in thalassemia were lower than in cases of iron-deficiency (p=0.023). In post-haemorrhagic anaemia, a significant increase in IRF and HYPO was observed when compared with healthy individuals (p< 0.001). In aplastic anaemia we found high values in the following parameters: IRF, MACRO, MCHr, HIPO, MRV and FEC. Higher values of MCHr, MRV and FEC were found in aplastic anaemia when compared to post-haemorrhagic anaemia (p-values of 0.009, 0.028 and 0.04, respectively). In patients with megaloblastic anaemia, we observed higher IRF, MACRO, MCHr, HYPO, MRV and FEC values than in healthy controls (p< 0.001). MACRO, MRV and MCHr values were higher in megaloblastic than in microcytic anaemia (p< 0.001). Considering the combination of several parameters, the accuracies on the prediction of different diagnoses were >91% (Table 1).
Conclusions: The development of new analysers implies the innovation of new parameters that provide added value for the diagnosis of different conditions. This study demonstrates the usefulness of new parameters provided by Mindray BC-6800Plus in the differential diagnosis of patients with anaemia.

P110
Clinical Utility Of Abbott Alinity Hq Extended Red Blood Cell Parameters In Differentiating Β-Thalassemia Trait And Iron Deficiency Anemia
Loredana Masi1, Sabatino Valente1, Fabrizio Papa1, Gabriella Lakos2, Zainab Mukhtar2
1Clinical Pathology Department, Rome, Italy, 2Abbott Diagnostics, Santa Clara, CA, United States

Introduction: Initial laboratory workup of microcytic anemias due to iron deficiency and β-thalassemia carrier status involves red blood cell analysis. Several mathematical formulas derived from red cell indices have also been suggested to aid in differentiation. The literature presents variable performance results possibly due to study population or analytical technologies. Advancements in technologies, algorithms and automation has led to the extended analysis of red blood cells. The objective of this study was to evaluate the performance of the Abbott Alinity hq advanced MAPPSTM-based RBC optical technology for the differentiation between Iron Deficiency Anemia (IDA) and β-thalassemia carrier status (β-thal).
Methods: Routine blood samples were tested in the laboratory of Fatebenefratelli Hospital (Rome, Italy) on Alinity hq. Among the 464 samples tested, 228 were identified as healthy controls, 30 as β-thal and 40 as IDA. ROC analysis was performed to evaluate the performance of red cell parameters showing statistically significant differences between the two microcytic anemia groups and of frequently used mathematical formulas
Results: RBC concentration was the most efficient discriminant (AUC of 0.963, Youden Index [YI] of 0.88) showing similar performance as RDW-SD (AUC of 0.960 and YI of 0.86), which outperformed RDW-CV (AUC of 0.924, and YI of 0.74). The absolute reticulocyte concentration also showed good diagnostic efficiency, with AUC of 0.808, and sensitivity of 79.3% and specificity of 74.4%. Hemoglobin distribution width (HDW), which is the %CV of the directly measured cellular hemoglobin concentration distribution, and CHCr, the average hemoglobin concentration of reticulocytes have emerged as novel discriminating parameters, with AUC of 0.749 and 0.785, respectively. The England & Fraser index was the best discriminating mathematical formula based on YI of 0.91, and the Ricerca, RDWI, Green & King and Mentzner Index formulas also showed strong discriminative power. The Shine & Lal index demonstrated moderate performance (sensitivity=83.0% and specificity=74.4%). The recent mathematical formula M/H, which is based on the percentage of microcytic and hypochromic red blood cells %Micro/%Hypo also showed moderate performance. (Table 1)
Conclusions: Extended red cell analysis delivered by the optical technology on the Alinity hq hematology analyzer is efficient for the initial discrimination of the two most common microcytic anemias (IDA and β-thal) in the laboratory.

P111
The Analysis Of Thalassemia Blood By Bc 6200
Thongperm Munkongdee, Nattrika Buasuwan, Suthat Fucharoen, Kittiphong Paiboonsukwong
Thalassemia Research Center, Nakhon Pathom, Thailand

Introduction:              Thalassemia is a very common genetic disorder worldwide.  Population-based screening, genetic counseling and prenatal diagnosis are essential to characterize couples at risk of conceiving fetuses with severe thalassemia for prevention and control of this disease.  An accurate and rapid analysis of blood is important for diagnosis.  We evaluated the analytical performance of the module Mindray BC-6200 for cytometric analysis of blood collected from different thalassemia disorders such as alpha and beta thalassemia carriers and common thalassemic diseases in Thailand and Southeast Asia.
Methods:        Total 63 cases of 9 thalassemia genotypes, the hematological parameters of white blood cells number (WBC), red blood cell number (RBC), RBC indices, nucleated red blood cells (NRBC) and platelet cell number (Plt) were measured by using Mindray BC-6200.
Results:       The RBC (mean ± SD) was observed with number 5.11 ± 0.50, 4.69 ± 0.58, 5.80 ± 0.80, 5.47 ± 0.47, 5.26 ± 1.21, 6.40 ± 0.70, 4.40 ± 1.00, 5.32 ± 0.84 and 3.49 ± 0.64 in normal, alpha-thalassemia 1 heterozygote, beta-thalassemia heterozygote, Hb E heterozygote, Hb E heterozygote with alpha-thalassemia, Hb E homozygote, Hb H-Constant Spring disease, Hb H disease and beta-thalassemia/Hb E disease, respectively.  The hemoglobin level (Hb) was demonstrated with low level in thalassemia disease; 15.3 ± 1.58, 10.8 ± 1.6, 11.9 ± 1.8, 14.1 ± 1.7, 11.3 ± 3.3, 12.9 ± 0.7, 8.5 ± 1.7, 9.4 ± 1.1 and 7.0 ± 0.9 in normal, alpha-thalassemia 1 heterozygote, beta-thalassemia heterozygote, Hb E heterozygote, Hb E heterozygote with alpha-thalassemia, Hb E homozygote, Hb H-Constant Spring disease, Hb H disease and beta-thalassemia/Hb E disease, respectively.  However, mean corpuscular volume (MCV) and mean corpuscular hemoglobin (MCH) were shown the decreased level in carrier and disease.  In case of splenectomy patients were presented an increased number of Plt and NRBC.  
Conclusions:       This preliminary results show that the BC-6200 can be used as a primary screening for thalassemia carriers and benefit in the study of thalassemic blood.

P112
Rdw-Cv By Alinity Hq, But Not By Sysmex Xn Is A Strong Differentiator Between Thalassemia And Iron Deficiency Anemia
Fabrizio Papa1, Loredana Masi1, Sabatino Valente1, Zainab Mukhtar2, Gabriella Lakos2
1Abbott Hematology, Santa Clara, CA, United States, 2San Giovanni Calibita Fatebenefratelli Hospital, Rome, , Italy

Introduction: The performance of red cell parameter-based discriminative formulae for the differentiation between thalassemia trait and iron deficiency anemia is inconsistent in the literature. One of the factors that has been suggested to contribute to the varying performance is the type and technology of hematology analyzers.
In this study we compared the performance of RDW-CV, RDW-SD and three commonly used mathematical formulas that include RDW in the equation, using results from the Abbott Alinity hq and the Sysmex XN-3000 analyzers for the discrimination of iron deficiency anemia (IDA) and beta-thalassemia trait (BTT).
Methods: Leftover blood samples were tested with Alinity hq and Sysmex XN-3000. Among the 464 samples that had results by both analyzers, 29 patients were identified as having BTT, 37 had IDA, and 200 were apparently healthy subjects. The discriminative function of RBC parameters and mathematical formulae was assessed by ROC analysis.
Results: RDW-SD was a strong performer with AUC of 0.958 and 0.932 by Alinity hq and Sysmex XN, respectively. RDW-CV by Alinity hq demonstrated strong discriminative power between BTT and IDA (AUC=0.924), however Sysmex XN was unable to distinguish between these conditions (AUC=0.503). Expressing RDW-CV in absolute units (mathematical SD: RDW-MATH SD), as suggested by some recent publications, improved the Sysmex XN AUC, but it still remained substantially lower than that for Alinity hq, as shown in the Table. The Alinity hq RDW-CV was higher in IDA (mean [SD]: 17.1 [2.3]) compared to BTT (13.6 [0.9]), and to the control group (13.2 [0.9]). The Sysmex XN RDW-CV, on the other hand, was not different between IDA and BTT (17.6 [2.6] vs 17.6 [1.5]) but was significantly higher in both patient groups than in the control population (13.3 [1.0]). Two of the discriminative indices demonstrated similar performance by the two analyzers, but the Ricerca index showed better discriminative power when determined using Alinity hq parameters.  
Conclusions: Our results have confirmed that the technology of hematology analyzers is an important contributing factor to the discriminative power of certain RBC parameters and mathematical formulas and demonstrated that the Alinity hq RDW-CV (along with RDW-SD) is a strong differentiator between BTT and IDA.

P113
Iron Deficiency Diagnostic Biomarkers Reporting: The State Of The Art In Italy
Fabrizio Papa, Anna Maria Cenci, Valentino Miconi, Barbara Casolari, Bruno Biasioli, Maria Golato
Italian Society of Pathology and Laboratory Medicine - Hematology Working Group (SIPMeL-GdSE), Castelfranco Veneto (TV), Italy

Introduction: The diagnosis, monitoring and managing of iron deficiency and iron deficiency anemia is usually based on several different hematologic and biochemical markers. In these conditions, it is currently debated the unavailability of complete performing tests in such important and widespread pathologies. Hematology Working Group of the Italian Society of Laboratory Medicine (SIPMeL-GdSE) investigated the use, standardization, diffusion and current reporting of the test concerning iron status by a survey about different iron biomarkers.  
Methods: SIPMeL-GdSE evaluated published data and recommendations from literature about this item and created a survey based on 27 questions concerning the analytical and reporting methods of the biomarkers involved in the iron deficiency definition. The survey was opened on the SIPMeL website for three months; the data were analyzed to evaluate practices and reporting of iron diagnostics in Italian laboratories. The questionnaires came from 57 laboratories of several geographical areas all over the country, different in structure, organization, test critical mass, so characterized: 54% (most of them) HUB Laboratories in public, private and university health services, high test number performing and large test panel offering; 28% CoreLab Laboratories, also with specialistic sections; 18% Spoke facilities.
Results: Ferritin and transferrin saturation test (less frequently) systematically used for iron-deficiency diagnosis. Reticulocytes counts and indices usually required by few specialist departments although widely available. Soluble transferrin receptor carried out by few laboratories (21.4%); the test can be sent to the HUB laboratories, but  low demand and availability is documented. 78.2% laboratories use chemiluminescence methods (CLIA) for ferritin quantification. Ferritin reporting includes reference values by sex in 89.29% and by age groups in 37.5%. Few laboratories reports (8.92%) offer decision making limits or thresholds, important  for the clinical interpretation of the result in different pathology areas (inflammation, comorbidities, age-related life conditions). Few laboratories (12.5%) use interpretative comments in reporting iron biomarkers.
Conclusions: The SIPMeL-GdSE survey highlighthed the need to improve tests and report standardization for iron deficiency diagnosis. This means to underline informative content and to introduce decision-making thresholds as recommended by scientific societies, to ensure correct interpretation and subsequent appropriate test based clinical decisions. To induce virtuous professional behavior, SIPMeL-GdSE introduced these items in the 2021 educational training-plan.

P114
Reticulocyte Haemoglobin Concentration (Chr) And Hypochromic Erythrocytes (%Hypo) Percentage In Screening Test For Iron Deficiency Anaemia In Cancer Patients
Loan Tran Thi Anh1,3, Tung Tran Thanh1, Suzanne MCB Thanh Thanh1,2, Ha Hoang Thi Thuy1, Vinh Tran Thanh1, Nhan Nguyen Tran Thien1, Anh Nguyen Thi Truc2, Tu Nguyen1
1Cho Ray Hospital, Ho Chi Minh, Vietnam, 2University of Medicine and Pharmacy at Ho Chi Minh City, Ho Chi Minh, Vietnam, 3Nguyen Tat Thanh University, Ho Chi Minh, Vietnam

Introduction: The early detection of iron deficiency anemia (IDA) can significantly enhance the treatment effectiveness as well as improve the living standard for cancer patients. The CHr can provide information about the bone marrow’s erythrocyte production status and response to iron therapy, and the % HYPO has also reflected the iron status during the last several months prior to clinical manifestation. This study is aimed to evaluate the possibility of CHr and % HYPO in screening IDA among the cancer patientsas a low-cost solution that has not been performed in Vietnam in order to early screen IDA for cancer patients with anemia through basic tests such as erythrocytes and reticulocytes. This study is to survey the Receiver Operating Curve (ROC) and the Area underthe Curve (AUC) to find out the use value of CHr and % Hypo. The Youden index is used to determine the optimal threshold value. 
Methods: A retrospective cross-sectional descriptive study of 347 participants, including healthy people (178), patients with IDA (39) and cancer patients with anemia (130), was conducted at Cho Ray Hospital.
Results: Screening IDA in the cancer patients based on CHr and % Hypo was observed with the combination of the two testing parameters of CHr and %HYPO that made an increase in distinguishing between IDA and non-IDA among the cancer patients with p< 0.05 that has AUC=0.686; 95% CI: 0.592-0.780 with p< 0.001; corresponding to the predicted values of CHr ≤29pg (Se:46%; Sp:74%) and %HYPO ≥8 (Se:50%; Sp:76%) that are recommended possible to optimize the screening for the IDA cancer patients, in which, the AUC of % HYPO was observed with poor performance when being used independently, but with an increased reliability when being combined with the CHr. The differentiation among the pairs of study group is significant with p< 0.003.
Conclusions: As with the low cost, facilitation and help to clinical diagnosis, the CHr and %HYPO appears to be a useful tool that is able to utilize for screening IDA for the cancer patients among the population of cancer patients with anemia.

P115
Investigation Of Gas Transport Function In Pregnant State With Covid-19
Ekaterina Yastrebova1, Elena Serebrennikova2
1Voevodsky Institute of Chemical Kinetics, Novosibirsk, Russia, 2Medical Centre of Siberian Branch of the Russian Academy of Science, Novosibirsk, Russia

Introduction: Erythrocytes play the leading role in gas transport function and are completely dedicated to their job of transporting respiratory gases (oxygen and carbon dioxide). First of all, the exchange diffusion of chloride and bicarbonate through erythrocyte membrane is a rate-limiting step for the transfer of CO2. It is carried out by Band 3 proteins which is placed in erythrocyte membrane. Detailed characterization of morpho-functional erythrocyte parameters is necessary to control risk of hypoxic stage in organism. This study tries to reveal erythrocyte parameters which deviations may appear during COVID-19 infection.
Methods: Whole blood was taken by venopuncture. Then cells were used in experiments at room temperature (22 °C). All donors were in pregnant state and divided in three groups: patients with severe and mild course of the disease induced by COVID-19 and healthy donors. Experiments were carried out on scanning flow cytometer (SFC). To enable characterization erythrocyte gas transport function we used theoretical simulation of light scattering profiles (LSPs). We obtained for each measured erythrocyte in samples the content of hemoglobin simultaneously with morphological characteristics: sphericity index, surface, volume. We measured these parameters in three states: native RBC, during lysis in ammonium chloride and after RBC activation by addition of nifedipine solution. We define the anion permeability in terms of mean number of ‘effective’ and ‘activated’ erythrocyte Band 3 proteins.
Results: Total 4 patients and 20 donors were included in this study. It turns out that patients with severe course of the disease induced by COVID-19 have decreased rate of ‘activated’ and ‘effective’ anion exchangers compare to healthy group and even to mild course group. Patients in both cases have the same ratio between ‘activated’ and ‘effective’ anion exchangers. It means that RBCs have capacity to increase the rate of anion exchange (we measured it by activation in vitro), but in vivo cells couldn’t satisfy the demand and it is result in hypoxic state. 
Conclusions: We observed the alterations in erythrocyte anion permeability during COVID-19 in pregnant state. In our assumption, there are mechanisms that indirectly affect the CO2/O2 exchange of erythrocytes in COVID-19 case. Result of this work can be used in the detection of new molecular-kinetics models and in monitoring of the anion rate exchange throughout the treatment procedure.

P116
High FastIng Blood Glucose Related Lgals3 Gene Was Connected With The Progression Of Chronic Myeloid Leukemia By Loss Of INtron 3
cui zhao, Shuqi Li, Jing Liu, Fangyi Yao, Yanmei Xu, Jing Zhang, Jin Lin, Tingyu Qin, Fangmin Zhong, Meiyong Li, Bo Huang, Xiaozhong Wang
Department of Clinical Laboratory, The Second Affiliated Hospital of Nanchang University, jiangxi , China

Introduction: Although chronic myelogenous leukemia (CML) has been effectively treated with the application of tyrosine kinase inhibitors (TKIs), CML patients in blastic phase (BP-CML) have gained little benefit from them. Hence, our objective was that exploring factors related to the progression of CML.
Methods: We enrolled eligible 304 patients with chronic phase-CML (CP-CML) and 117 BP-CML patients in this retrospective study. Fasting blood glucose (FBG) level and LGALS3 gene and its noval splicing variant expression were compared in BP-CML and CP-CML by receiver-operating characteristics (ROC) curve analysis, propensity score matching (PSM) and real-time quantitative PCR (qPCR).
Results: FBG level in CP-CML and BP-CML were 3.74 ± 1.52mmol/L and 5.81 ± 1.94 mmol/L, respectively. The level of FBG in BP-CML was significantly higher than that in CP-CML (P 

Best Student Presentation Award Finalist
P117
Effectiveness Of Simple Equations In Screening Spurious Hemolysis In Whole Blood Specimen
Nouha Ben Amara , Aya Chakroun, Rania Hadj Taieb, Nesrine Zmerli, Amani Jabri, Raouaa Ismail, Sonia Mahjoub, Hela Baccouche, Neila Ben Romdhane
La Rabta Hospital, Tunis, Tunisia

Introduction: Hemolyzed samples may produce unreliable results of complete blood count (CBC). So detecting hemolysis in vitro is a challenge in our quality approach . We therefore proposed to evaluate the performance of two discriminant formulas developed by Lippi and al for the detection of hemolysis in vitro.
Methods: Peripheral blood samples were collected in anticoagulant K2-EDTA . First CBC analysis was realized on Sysmex XT4000i hematology analyzer then, visual assessment of hemolysis was realized and the hemolyzed samples were excluded .Thus 217 samples were enrolled (H0). In a second step, all specimen were subjected to mechanical hemolysis according to the Dimeski method: The anticoagulated total blood of the second aliquot was passed 5 times(H1 mild hemolysis) then 10 times (High hemolysis)  through a small caliber needle. After each hemolysis, a CBC was performed. All the data were analyzed on SPSS 19.0 software. Receiver operative characteristic curve with area under the curve (AUC), sensitivity, specificity, accuracy  were determined to assess performance of  F1 and F2 (formula 1(F1) : [Ht/Hb]x100 and formula 2(F2) : [Ht/Hb] x √MCV).
Results: Standards CBC parameters as well as erythrocyte research parameters showed a statistically significant difference for the samples with a high degree of hemolysis in comparison with the sample group without hemolysis. Platelet count, PDW, PCT, P-LCR and all RBC parameters were altered by the spurious hemolysis  . In multivariate analysis, both formulas showed poor discriminating power to identify hemolyzed specimen (AUC = 0.41 with F1, AUC= 0.39 with F2), a very poor sensitivity (30.6% and 20.5% respectively for F1 and F2) and only allowed an accuracy < 50%. Furthermore, the degree of agreement between the visual identification of hemolysis and that attested by each Lippi’s formula was poor ( Kappa index= 0,02).
Conclusions: This study showed the interference of in vitro hemolysis on the blood cell-counting parameters performed by the impedance method, as well as search parameters. Moreover as far as we know, no data have been reported in using Lippi’s Formulas on detecting hemolysis. Our results did not show the reliability of these formulas.

P118
Infographics As A Valuable Tool For Disseminating Recommendations: The Case Of Plasmacellular Disorders Management
Anna Maria Cenci1, Bruno Biasioli1, Barbara Casolari1, Valentino Miconi1, Fabrizio Papa1, Ignazio Brusca2, Maria Golato1,3
1Italian Society of Pathology and Laboratory Medicine - Hematology Working Group (SIPMeL-GdSE), Castelfranco Veneto (TV), Italy, 2Italian Society of Pathology and Laboratory Medicine - Proteins Working Group (SIPMeL-GdSP), Castelfranco Veneto (TV), Italy, 3Italian Society of Pathology and Laboratory Medicine - Oncology Working Group (SIPMeL-GdSDO), Castelfranco Veneto (TV), Italy

Introduction: Infographic is a modern tool that organizes information in a visual way by minimizing the text part. It is the newest and simplest form of data visualization, and tries to explain and communicate complex concepts clearly and quickly. The textual elements are short, often with a didactic function. The infographic allows for an "at a glance" narrative, and necessarily needs of a skilled teamwork for successful results.
Methods: In 2018, the SIPMeL-GdS-Proteins published the recommendations on the laboratory paths of Monoclonal Gammapathies. To facilitate dissemination, the infographics seemed the most suitable tool. In fact, the following points seemed particularly fitting to this tool: the presence of an appropriate "Story telling", well-identified subject and recipients; clear and consolidated clinical laboratory paths; an important diagnostic proposal based on technological innovations; several requests for clarification/standardization coming for a long time from laboratory professionals. It was also possible to get collaboration between several SIPMeL study groups for their specific competences (GdS-Hematology; GdS-Proteins; GdS-Oncology) and there was the opportunity to simultaneously address medical specialties and the laboratory world.
Results: The realization employed about two months; the work was followed up with face-to-face and web-meetings for the control and implementation of the work phases (project; graphic hypothesis, meeting with graphic designers, proof-reading; printing). The infographic was presented during the first session of the 2019 National Conference in Riva del Garda, and distributed to the meeting attendees.
Conclusions: The experiment was very successful and appreciated. Nowadays the distribution to clinicians and general practitioners, often previous problematic requested, is ongoing so the SIPMeL website publication. In our experience, the infographic results an essential simplification of complex concepts, and not a oversimplified realization. This new and effective way of promoting, distributing and advertising scientific activities seems to be particularly suited to the dissemination of recommendations and guidelines. On the basis of the previous experience, a second infographic, about the iron deficiency anemia diagnostic pathways, was therefore planned and implemented by the GdSE in 2020, currently awaiting printing.

P119
The Weakest Link: Results Of An Audit Into Patient And Specimen Identification Practices For Cbc Testing In A Pediatric And Maternity Hospital
Jason Ford1, Dana Al Eshaq1, Daniel Gebrekidan2, Eileen McBride1
1Sidra Medical and Research Center, Doha, Qatar, 2Health Partners, Eden Prairie, MN, United States

Introduction: There is abundant evidence which supports the use of barcodes in clinical laboratory practice, but there is less evidence regarding real world barcode compliance. Even in a fully barcoded hospital environment, misidentification may still occur: for example, if the blood collector bypasses the barcode system and instead manually enters information into the Laboratory Information System (LIS). After an episode of patient/sample misidentification, we conducted an audit of barcode scanning compliance in the collection of Complete Blood Count (CBC) samples in a tertiary care pediatric and maternity hospital.
Methods: All CBC patient and sample barcode results were assessed for the period 1 January 2020 to 31 December 2020, using data retrieved from the hospital LIS. We were able to evaluate barcoding compliance with Positive Patient Identification (i.e. PPID) and Positive Accessioning Identification (i.e. specimen identification, PAID) throughout the year.
Results: In 2020, 46272 unique CBC accession numbers were generated in 19057 patients. These were collected on patients 0-< 6 months (13.6%), 6 months to < 2 years (10.0%), 2-< 12 years (36.8%), 12-< 18 years (18.5%), and >18 years (21.1%). Overall, 68% were collected with proper PPID, and 82% were collected with proper PAID, leaving 32% PPID noncompliance and 18% PAID noncompliance. Rates of noncompliance by patient age are shown in Table 1.
Table 1. PPID and PAID noncompliance by patient age.
Patient age PPID noncompliance (%) PAID noncompliance (%)
0 - < 6 months 28.2% 9.7%
6 months – < 2 years 33.7% 15.5%
2- < 12 years 29.6% 13.8%
12-< 18 years 29.9% 15.7%
>18 Years 39.0% 33.2%

Conclusions: Our audit has uncovered an instance of poor barcode scanning compliance in a pediatric and maternity hospital. These patients are at higher risk of misidentification, which could complicate diagnosis and management. This highlights that the existence of a robust barcode scanning system does not obviate the need for careful staff training and regular audits to support compliance.

P120
The Problem Of Choice: Excess Configurability Of The Hospital Information System And Its Consequences In The Clinical Hematology Laboratory
Jason Ford, Wissam Khaiwi, Elicia Ginn, Eileen McBride
Sidra Medical and Research Center, Doha, Qatar

Introduction: Numerous authors have recommended a variety of configurations and rules for laboratory and hospital information systems (LIS, HIS), to facilitate appropriate test ordering and reporting. Our new laboratory has made effective use of some of these configurations and rules as we have set up a new tertiary care pediatric hospital. We encountered an unexpected complication of extensive HIS configurability: more configurable systems are more challenging to set up, difficult to validate, and can make it difficult or impossible to apply rules.
Methods: The Department of Pathology began developing specific LIS and HIS procedures in 2014. Outpatients were first seen in 2016, and inpatients in 2017. We reviewed our experiences in developing and improving the LIS and HIS from 2014 through 2021, including validation of our test ordering and test reporting, and the establishment of rules designed to improve clinical communication and patient care.
Results: The LIS at this hospital was set up simultaneously but separately from the setting up of the clinical aspects of the HIS, as almost infinitely configurable systems. The LIS/HIS design approvals process was designed mainly to support clinical consensus rather than LIS/HIS coherence and simplicity. Ultimately the LIS and HIS were configured to accommodate the complex preferences of scores of physician experts. Our highly configurable system does in practice meet those preferences, but at the cost of system complexity. This has had numerous laboratory and clinical consequences, including: the separation of LIS and HIS building early on has led to challenges in test naming and ordering owing to different preferences on the part of clinicians and pathologists; laboratory test results are reported to the HIS in multiple different ways, which makes it extremely difficult to validate results reporting; diagnostic and procedural coding is extraordinarily complex, which makes the establishment of diagnosis-related rules nearly impossible.
Conclusions: Highly configurable hospital information systems are an excellent tool, but the more configurable the HIS the more complexity is introduced into the system. This can lead to confusion and can make it difficult or impossible to complete necessary validations or implement certain clinical rules. The degree of configurability of hospital information systems should be limited to permit safer clinical practice.

P121
Evaluation Of Various Aptt Reagents For Monitoring Jivi &Ndash; One Centre&Rsquo;S Experience
Karen A. Moffat1,2, Stephen A. Carlino1, Catherine P.M. Hayward1,2,3
1Hamilton Regional Laboratory Medicine Program, Hamilton, ON, Canada, 2Department of Medicine, McMaster University, Hamilton, ON, Canada, 3Department of Pathology and Molecular Medicine, McMaster University, Hamilton, ON, Canada

Introduction: Our laboratory supports a regional program for adult and pediatric hemophiliacs from a large geographical region. Recently, Canadian Blood Services approved JIVI, a recombinant B-domain deleted PEGylated FVIII concentrate, as an approved replacement therapy for Hemophilia A patients. Published studies report significant differences between APTT reagents in measured FVIII activity due to JIVI. Our goal was to compare our current to recommended APTT reagents for monitoring FVIII levels in hemophilia A patients receiving JIVI.
Methods: FVIII activity was evaluated with APTT reagents Dade Actin FS (current reagent), Dade Actin FSL (Siemens, Marburg, Germany); and HemosIL SynthASil and SynthAFax (Instrument Laboratories, Bedford, MA). Validations included assessments of linearity, two levels (normal (Control N) abnormal (Control P)) of quality controls (QC)(Siemens), normal plasmas (n=20; Affinity Biologicals, Ancaster, ON) and plasma from patients with hemophilia A receiving JIVI (n=37, tested with select reagents). Chromogenix COAMATIC FVIII (Instrument Laboratories, Bedford, MA) was used to verify recovery of FVIII activity from the ECAT BAY94 JIVI validation set (Voorschoten, the Netherlands). Statistical analysis included: mean, SD, %CV, regression, Student’s t-test (p< 0.05 considered significant).
Results: Both levels of FVIII QC were within established (Actin FS) and manufacturer specified ranges with all four APTT reagents. Similar to Actin FS, the other APTT reagents demonstrated acceptable FVIII linearity (0.02 to 0.80 U/mL), and transferable RI (0.50–1.50 U/mL). Chromogenic FVIII levels for the ECAT BAY94 JIVI validation set were similar to FVIII levels determined with APTT reagents (Table 1). Further testing of samples from hemophilia A patients receiving JIVI indicated that their plasma FVIII activity measured with Actin FS significantly overestimated the FVIII activity, compared to SynthAFax (p< 0.001).
Conclusions: Our study confirms that Actin FS over-estimates FVIII activity due to JIVI. We verified that APTT-based FVIII assays using Actin FSL, SynthASil, or SynthAFax measured appropriate FVIII levels in assessments of the ECAT BAY94 evaluation set. FVIII levels evaluated with Actin FSL, SynthASil, or SynthAFax showed similar performance in evaluations of QC samples, linearity and RI transference. Our laboratory has now implemented an APTT-based FVIII assay using SynthAFax for monitoring FVIII levels in samples from patients with hemophilia A receiving JIVI replacement.

Best Student Presentation Award Finalist
P122
Hematology References Range Of Adult Subject In Indonesia Based On Gender And Age Group According To Clsi Guidelines
Tri Ratnaningsih1, Rahadian Faisal1, Afifah Khairu Nisa2, Usi Sukorini1
1Universitas Gadjah Mada, Yogyakarta, Indonesia, 2Universitas Ahmad Dahlan, Yogyakarta, Indonesia

Introduction: The reference range of hematologic parameters is essential for clinical decision making. The objective of this study is to establish the reference range for Indonesian adult subjects for hematology.
Methods: A cross-sectional observational study was determining reference ranges based on CLSI EP28-A3c. Research in Yogyakarta in February 2019-July 2020. A total of 2 mL of blood was collected by phlebotomy and collected in a K2-EDTA tube then examined using the ADVIA 2120 hematology analyzer. Data analysis used MedCalc 13.0.6.0. Subjects were divided based on gender and age group. Outlier data were tested using Tukey. Kolmogorov-Smirnov tested the normality of data distribution. Different from the mean using independent t-test or Mann-Whitney. The establishment of reference ranges used the non-parametric percentile method and the robust method.
Results: One thousand three hundred seventy subjects consisted of 536 (39.1%) male subjects and 834 (60.9%) female subjects. Based on age, it was divided into 1255 (91.6%) young adult subjects (18-34 years) and 115 (8.4%) old adult subjects (35-59 years). Men had significantly higher scores than women on erythrocyte, hemoglobin, hematocrit, MCH, MCHC, PDW, and monocyte parameters, and lower on platelet parameters. Young adults had a significantly higher number than older adults on leukocytes' parameters, erythrocytes, hemoglobin, hematocrit, MCV, lymphocytes, and basophils. The reference ranges obtained in this study are representative of the healthy adult population in Indonesia. Based on physiology, gender is associated with various factors such as hormones androgens, estrogen, and testosterone on hemopoiesis.
Conclusions: This study determined the reference range for the number of leukocytes, erythrocytes and platelets, hemoglobin, hematocrit, erythrocyte and platelet indexes, and the leukocyte differential count in Indonesian healthy adult subjects using the ADVIA 2120 hematology analyzer.

P123
Establishment Of Novel Complete Blood Count Reference Values For Healthy Pakistani Adults
Muhammad Shariq Shaikh1, Sibtain Ahmed1, Zeeshan Ahmed1, Attika Khalid2
1Aga Khan University Hospital, Karachi, Pakistan, 2Foundation University Medical College, Islamabad, Pakistan

Introduction: In laboratory medicine, the reference interval (RI) for quantitative anlalytes is a range of values that is deemed normal in a given population. These population based RIs are of significant importance for appropriate decision making regarding patient’s health. Hematological RIs, are in particular, markedly influenced by variety of factors. Therefore, accrediting bodies also mandate that each laboratory should establish its own RIs in its own population.
Methods: This prospective anlaysis was conducted at the Section of Hematology, Department of Pathology and Laboratory Medicine, the Aga Khan University Hospital (AKUH), Karachi, Pakistan. Twenty-two complete blood count parameters were measured in adults aged 18-60 years who passed the initial health screening questionnaire. All samples were handled strictly following standard operating procedures. Microsoft Excel and EP Evaluator software were used for statistical analysis. Nonparametric CLSI EP28-A3C method was used to establish upper and lower confidence limits at 90% significance.
Results: A total of 323 participants passed the questionnaire and short-listed for blood collection. There were 147 males and 176 females. Reference intervals were established in 297 participants after exclusion of 26 outliers with grossly abnormal test results. All the quality control values and instrument tolerance limits were acceptable before sample analysis. Analytes included: red cell count, hemoglobin, hematocrit, mean cell volume, mean corpuscular hemoglobin, mean corpuscular hemoglobin concentration, red cell distribution width, reticulocyte count, white cell count, percentages and absolute values of neutrophils, lymphocytes, monocytes, eosinophils, basophils, neutrophil to lymphocyte ratio, platelet count and immature platelet fraction. White blood cell count and RBC indices were calculated separately in male and female cohorts.
Conclusions: Complete blood count RIs established in this study showed significant differences from RIs in Caucasian population. Each hematology laboratory should establish RIs in population it serves in order to have valid interpretation of laboratoy results.

P124
Proposal For An Optimized Protocol To In Vitro Red Blood Cell Glycation
Marcus Silva1, Analia Alet1, Bibiana Riquelme1,2
1Facultad de Cs. Bioquímicas y Farmacéuticas (UNR), Rosario, Argentina, 2Facultad de Cs. Bioquímicas y Farmacéuticas (UNR), Rosario, Argentina, 3Grupo de Física Biomédica, IFIR (CONICET-UNR), Rosario, Argentina

Introduction: Glycation is a term that refers to slow, non-enzymatic, and spontaneous reactions between the amino groups of proteins and the carbonyl group of a reduction sugar, like glucose. In pathological conditions, such as diabetes, due to the exposure time of proteins and the level of concentration of the glucose in the blood, advanced glycation end-products are observed. With the aim to optimize an in vitro model for the glycation phenomenon, some of the most relevant in vitro glycation protocols found in the literature were analyzed to compare them with each other. Through this analysis, an optimized protocol for this type of study is proposed.
Methods: After analyzing the available literature, a table was created to summarize the most important details about the parameters for glycation. The parameters as incubation medium, temperature and time, hematocrit, and glucose concentration were compared to establish the optimal condition.
Results: Most of the hematocrit values vary from one bibliographic reference to another. In general, the hematocrit does not exceed the value of 50%. The most commonly used incubation media are phosphate-buffered saline (PBS). However, it was also found cell culture media Dulbecco’s modified Eagle’s medium (DMEM) and antibiotics like penicillin and streptomycin. All authors used the temperature of 37 °C, which is the average body temperature. The values of concentration of glucose during the incubation, found in the consulted literature, vary between 0.09 g/dL and 6 g/dL. Incubation times that vary from 0.5 to 120 hours are used. Also, different techniques were used for the analysis of the effects of glucose on RBC: spectrophotometry, mass spectrometry, cell transit time analysis, glucometer, Erythrocyte Rheometer, digital image analysis, and Optical Chip Aggregometer.
Conclusions: The parameters can be variable according to the researcher's objective. However, it should be considered that to achieve the glycation effect, the glucose values must be higher than those found under normal conditions. We suggest a range between 0.1 g/dL and 0.3 g/dL. The incubation media recommended is PBS, and 6 hours of incubation to avoid contamination. The hematocrit during the incubation should be as close to the physiological values, being the most appropriate value of 40% (v/v).

P125
Extended Stability Study Of Blood Count Parameters
Teresa Villalba, Jorge Medina, Laura Redondo
Catlab, Viladecavalls, Spain

Introduction: Blood count analysis is one of the most requested tests in the laboratory. Blood count parameters also vary in the patient in various clinical situations. In certain situations we have to analyze samples extracted more than 24 hours ago (for results verification, studies added in pretransfusion samples, for example) and we wanted to know what the variation of results is over time and whether the results we obtain in this deferred analysis can be considered reliable. 
Methods: We selected 20 samples with results in different ranges of values analyzed in the first 4 hours after extraction.  The analysis was repeated at 10, 30, 54, 78, 102 hours after extraction and even a determination was made at 7 days. The first 12 hours the samples were kept at room temperature and then stored at 4oC until 30 minutes before each analysis. The samples were analyzed in a Sysmex XN20 analyzer. The byass of each result was analyzed against the initial result. For reference, the maximum total error specifications were used for each magnitude accepted in our laboratory.
Results: The different results of the 20 samples were recorded. Here we show (Table 1) the byass average for each magnitude at defined analysis times. 
MAGNITUDE/SPECIFICATION% 31h 55h 79h 103h 175h
WBC/15,5 1.35 0.64 -0.75 -1.94 -5.05
RBC/4,4 -0.26 -0.93 -1.17 -1.09 -1.7
HB/4,2 0,39 0,28 0,26 -0,12 0,56
HTO/5,9 1,93 2,69 3,36 5,39 8,67
MCV/3,6 2,19 3,65 4,60 6,54 10,54
PLT/16 -1,05 -1,98 -3,83 -3,36 -6,46
NEUT/23,4 3,82 2,14 0,25 -1,20 -2,51
LYMPH/17,6 -1,72 -1,32 -2,23 -3,15 -6,05
MONOC/27,6 2,70 3,49 3,81 2,40 -11,42
RETIC/16,8 0,11 -1,65 -2,74 -4,38 -3,55
The specifications sources are the biologic variability database from SEQC  and analitical spanish societies consensus. From our results we concluded the majority of the magnitudes are stable almost five days except MCV (only two days) and Hematocrit (only four days).  We reccomend use this data as a frame to know the average byass of the results butdon't delay the samples analysis.  Morphology studies are not reliable after 12-24 hours.  Further studies are needed to evaluate more accurately these data.
Conclusions: - RBC and Hemoglobin values are stable almost five days stored a 4º C. - Hematocrit and MCV values only are stable four and two days at the same conditions. - Further studies are needed to confirm these data

P126
Initial Assessment Of Malutilization Of Iron Status Testing In Ottawa, Canada And Recommendations For Improved Utilization
Qian Wang1, George Cembrowski2,3, Yuelin Qiu4, Christopher McCudden5
1Department of Biochemistry, University of Alberta, Edmonton, AB, Canada, 2Department of Laboratory Medicine and Pathology, University of Alberta, Edmonton, AB, Canada, 3Cembrowski-Cembrowski Quality Control Consulting, Edmonton, AB, Canada, 4Faculty of Engineering, University of Alberta, Edmonton, AB, Canada, 5Department of Pathology and Laboratory Medicine, Division of Biochemistry, University of Ottawa, Ottawa, ON, Canada

Introduction: The laboratory diagnosis of reduced or increased iron stores is problematic as iron, TIBC, and ferritin levels are influenced by chronic disease and inflammatory disorders.  We examine the last 5 years of Ottawa Hospital serum/plasma iron, TIBC, ferritin and transferrin utilization data to delineate (in)appropriate iron utilization, concentrating on the addition of transferrin to total iron binding capacity (TIBC) and iron results.
Methods: We obtained five years (2014-2019) of deidentified Ottawa Hospital inpatient and outpatient ferritin, iron, TIBC, transferrin, date and time of their sampling and patient age. Same day tests were identified and grouped into discrete test sets and presented as a Venn diagram which enabled the comparison and evaluation of the different patterns of iron test utilization. Linear regression was used to demonstrate the equivalence of transferrin and TIBC.  We tested our hypothesis that transferrin and TIBC were used interchangeably when ordered with ferritin by comparing the relative cumulative frequency distributions of ferritin levels when ordered 1) either separately or either with 2) transferrin and iron or 3) TIBC and iron or 4) TIBC and iron and transferrin.  
Results: The Venn diagram (figure) shows the motely test assortments used to diagnose iron status. Linear regression demonstrated that transferrin level can be determined from TIBC with high accuracy. 76.7% of ferritins were ordered alone, 23.3% were ordered in combination with other iron markers. Among these, 18% of ferritin is ordered in combination with TIBC and iron. The ferritin frequency distribution of ordered alone, or in combination with other iron markers were very similar.  
Conclusions: In laboratories providing TIBC and iron testing, transferrin offers no advantage over TIBC and should not be ordered. In patients without inflammatory disorders or with low to moderate inflammation, ferritin should be used to assess iron status.  TIBC and iron are recommended for patients with high ferritins. To influence more efficient testing for iron status, Venn diagrams of iron utilization should be shared with the appropriate care groups.

P127
 Lactate Dehydrogenase-To-Monocyte Ratio And Lactate Dehydrogenase-To-Uric Acid Ratio Were Connected With Disease Progression In Cml Patients
cui zhao, yanmei xu , shuqi li, bo huang, xiaozhong wang
Jiangxi Province Key Laboratory of Laboratory Medicine, Department of Clinical Laboratory, The Second Affiliated Hospital of Nanchang University., jiangxi, China

Introduction: The study aimed to investigate the diagnostic roles of the lactate dehydrogenase-to-monocytes ratio (LMR) and the lactate dehydrogenase-to-uric acid ratio(LUR)with the disease progression of chronic myeloid leukemia ( CML).
Methods: A total of 207 CML patients, including 134 CML-chronic phase(CML-CP)patients and 73 CML-advance phase(CML-AD)patients, and 120 healthy controls were selected in this retrospective study. All clinical and laboratory parameters of the CML patients were extracted from their medical records. The correlations between LMR, LUR, and the clinical characteristics with the disease progression were analyzed. 
Results: For CML patients, the level of LUR was significantly higher than the healthy control group, the level of LMR was significantly lower than the healthy control group (P 

3:00 - 4:15 PMVirtual Room 2
PLENARY - PRESIDENTIAL SYMPOSIUM

Chair(s): Tracy George
3:00
Full-Field Hemocytometry. More Than Forty Years Of Progress Seen Through Clinical & Laboratory Haematology And The International Journal Of Laboratory Hematology (1979-2021)
Giuseppe d'Onofrio
Catholic University of Rome, Italy

The extraordinary advances in clinical hematology, biology, and oncology in the last decades would not have been possible without discovering how to identify and count the cells circulating in the blood. For centuries, scientists have used slides, counting chambers (hemocytometers), and diluting and staining solutions for this task. Then, automated hemocytometry began. This science, now linked to the daily routine of laboratory hematology, has completed an overwhelming path over a few decades. Our laboratories today operate with versatile multi-parameter systems, ranging from complex single-channel instruments to bulky continuous flow machines. In terms of clinical information obtained from a simple routine blood test, the full exploitation of their potential depends on the operators' imagination and courage. A comprehensive review of the scientific publications that have accompanied the development of hemocytometry from the 1950s to today would require entire volumes. More than seven hundred contributions that authors worldwide have published in Clinical and Laboratory Haematology until 2007 and then the International Journal of Laboratory Hematology are summarized.  Such journals have represented and hopefully will continue to represent the privileged place of welcome for future scientific research in hemocytometry. Improved technologies, attention to quality, new reagents and electronics, information technology, and scientist talent ensure a more profound and deeper knowledge of cell properties: current laboratory devices measure and count even minor immature or pathological cell subpopulations. Full-field hemocytometry includes the analysis of non-hematic fluids, digital adds to the microscope, and the development of effective point-of-care devices.


3:30
Pathology From The Subcellular Level On Up
Garry Nolan
Stanford University

I will present evidence of deep internal order in immune functionality demonstrating that differentiation and immune activities have evolved with a definable “shape”.  Further, specific cellular neighborhoods of immune cells are now definable with unique abilities to affect cellular phenotypes—and these neighborhoods alter in various cancer disease states.   In addition to cancer, these shapes and neighborhoods are altered during immune action and “imprinted” during, and after, pathogen attack, traumatic injury, or auto-immune disease.  Hierarchies of functionally defined trans-cellular modules are observed that can be used for mechanistic and clinical insights in cancer and immune therapies.