Abstracts

Laboratory Corporation of America

An unexpected PT and/or APTT prolongation is often evaluated using an invitro “mixing” test with normal plasma. Failure to correct (“non-correction”) prolongation upon mixing is attributed to an inhibitor, whereas “correction” points to factor deficiency(ies). There is, however, lack of standardization around what defines “correction” versus “non-correction” of a mixing test. For this reason, an international, multi-site study evaluating mixing tests using multiple PT and APTT reagents for the purpose of defining an optimal method for determining “correction” or “non-correction” using well-characterized plasma samples was performed. Mixing study results were evaluated using 11 different calculation methods. Misprediction, representing the failure of a mixing study interpretation method to correctly identify a factor deficiency versus inhibitor was used to assess the results. Study results revealed that percent correction and the Rosner Index are more suitable than other calculation methods for interpreting APTT and PT mixing test results for nearly all reagents evaluated. In general, other calculation methods that performed well in the identification of factor deficiency tended to have high misprediction rates for inhibitors and vice versa. No single method of mixing test result calculation was consistently successful in accurately distinguishing factor deficiency(ies) from inhibitor. Between reagent and site variability was also identified.

¹College of Medicine University of Lagos Lagos, Nigeria, ²General Hospital Odan-Lagos, Nigeria, ³Evercare Hospital Lekki Lekki- Lagos, Nigeria

Introduction: Authors reckon that clinically significant errors occur in laboratory practice. If not mitigated, laboratory errors can lead to missed/delayed diagnosis, improper monitoring of patients and poor clinical outcomes. There is a continual need to report and manage all forms of error in order to reduce them to nil or negligible levels through effective root cause analysis, corrective/preventive actions and process improvement strategies. The objective of the study is to evaluate the occurrence management system in a private tertiary hospital in Lagos Nigeria. In particular, types of incidences, individuals affected, laboratory phases/elements affected, root causes and incidence closure were described. Methods: Index study was conducted at a 165-bed private tertiary care hospital in Lagos, Nigeria. The hospital laboratory units include phlebotomy rooms, blood donor clinic, haematology, clinical chemistry, microbiology, histopathology, PCR laboratory and blood bank. Electronic reports of laboratory incidents submitted over a period of 18 months between April 2021 and September 2022 were retrieved from the logs, collated and analyzed. Data collected included types of incidences, location of incidence, individual affected, incidence summary, laboratory phase/element involved, reporting personnel, root causes, incidence closure and reasons for non-closure, if any. Descriptive data are presented as frequencies and proportions. Results: Of the 81 reported occurrences, there are 72 (88.9%) actual errors (adverse events, no harm events, and sentinel events), others were near misses (12.3%). Ten (12.3%) occurred outside the laboratory. In terms of potentials for harm, patients (77.8%) and staff (8.6%) were largely affected. Most errors involved the extra-analytic phases (43.2% pre-analytic and 19.8% post-analytic). Others errors were 13.6% intra-analytic, 11.1% safety events and 6.2% related to external examination (test referrals). Eight different categories of laboratory staff were involved in reporting incidences, with most reports generated by medical laboratory scientists (40.7%), phlebotomist (34.6%) and pathologists (7.4%). Root cause analysis revealed employee performance issue (64.2%) and lack of effective communication (11.1%) as the most important. Reported incidences were fully closed out (90.1%) of the time with implementation of CAPA. Overall, 97.5% of the reported errors are considered preventable. Conclusion: Perhaps, there is no such thing as an error-proof clinical laboratory. All clinical laboratories should ensure functional occurrence management system to capture, report, correct and prevent laboratory errors. Effective error management processes coupled with continual training for all laboratory users, automation of processes and continual vigilance will prevent large number of laboratory errors and promote better outcomes for the laboratory.

The Aga Khan University, Hospital Karachi, Pakistan

Introduction: Inheritedcoagulation bleeding disorders are characterized by the absence or reduced levels of clotting factors, and the manifestations vary according to the type and magnitude of the deficient factor. Clotting factor deficiencies may lead to serious hemorrhagic manifestations and even death, if not diagnosed timely and treated appropriately. Objective: To determine the frequency of different clotting factor deficiency in an academic medical center. Methods: This cross-sectional study was conducted at the section of Haematology & Transfusion Medicine, the Aga University hospital from June to September 2022. A total of 200 clotting factor assays were performed and included in the study. Blood samples were collected in sodium citrate tube and clotting factors were performed on automated coagulation analyzer through clotting method. A quality control check was performed before analysis of blood samples. Results: Out of 200 clotting factor assays performed, 98 were found to be normal and 102 were found to be clotting factor deficient. Seventy-one (69.6%) were male and thirty-one (30.4%) were female. The mean age was 28 years (5 days–74 years). Eighty-seven (85%) patients were younger than 18 years of age. In males, the most common clotting factor deficiency was factor VIII in 32 (45%), followed by deficiency of factor X in 12 (17%), factor VII and factor IX in 7 (10%) each, factor V in 5 (7%), factor I in 3 (4%), combined deficiency of factor V &VIII in 2 (3%) and factor XI, factor XII and factor XIII in 1 each (1%). Vitamin K dependent clotting factor deficiency was found in 1 (1%). In females, factor VIII deficiency secondary to von Willebrand disease was found in 9 (29%), followed by factor V and X deficiency in 6 (19%) each, factor VII in 4 (13%), factor XIII deficiency and combined deficiency of factor V & VIII in 2 (6%) each, factor I and factor XII deficiency in 1 (3%) each. Conclusion: Factor VIII deficiency was found to be the most common clotting factor deficiency in both males and females. In females factor VIII deficiency was secondary to von Willebrand disease.

Key words:Clotting factors, bleeding, vitamin K.

¹Clinical Pathology Department, Faculty of Medicine, Mansoura University Mansoura, Egypt, ²Department of Biomedical and Molecular Sciences, School of Medicine, Queen's University Kingston, ON, Canada, ³School of Baccalaureate Nursing, St Lawrence College, Kingston, ON, Canada, ⁴Oncology Department, Faculty of Medicine, Mansoura University Mansoura, Egypt

Background: Risk prediction of venous thromboembolism (VTE) is a compelling challenge in hematological malignancies. Current guidelines do not recommend thromboprophylaxis for primary prevention, but assessment of the patient's individual risk of VTE prior to chemotherapy using risk assessment models (RAM) is generally advocated. Efforts have been devoted to building accurate predictive tools such as Khorana and Vienna CATS score for VTE risk assessment in cancer patients but those have not been validated for hematological malignancy. We aimed to assess validity of the Vienna Cancer and Thrombosis Study (CATS) score in prediction of VTE in a variety of hematological malignancies. Methods: a prospective cohort study conducted on 81 newly diagnosed cancer patients undergoing chemotherapy. Demographic, clinical and cancer related data were collected including age, sex, BMI, VTE risk factors and associated comorbidities. Patients were followed up for 6 months for VTE events during chemotherapy. Khorana score was calculated. Plasma D-dimer and soluble P-selectin (sP-selectin) were measured then Vienna CATS score was calculated. We assessed modified Vienna CATS by using new cut off levels of d-dimer and sP-selectin based on ROC curve of the patients' results. Results: 81 patients (average age 42.6 years; 50% females) were assessed. A 2.7% had advanced cancer with metastasis. The most frequent cancer in this group was Non-Hodgkin lymphoma (39.5%) followed by multiple myeloma (14.8%). All patients completed follow up and only 8 (9.8%) developed VTE events. The calculated probability of VTE occurrence using Khorana, Vienna CATS and modified Vienna CATS scores at cut off levels ≥3 were 87.5%, 87.5%, 100% respectively. The AUC in ROC curve of modified Vienna CATS score showed significant difference when compared to that of Vienna CATS and Khorana scores (P= 0.047 and 0.029, respectively) Fig. 1. Conclusions: Our data shows the usefulness for application of three VTE risk assessment models in hematological malignancies. Modified Vienna CATS score is more specific in detection of hematological cancer patients at risk of VTE when compared with Vienna CATS and Khorana score, while all three scores have similar sensitivity. Implementation of RAM in clinical work up of cancer patients could help clinicians to tailor anticoagulant therapy and lead to improved use of thromboprophylaxis.

¹King Abdullah International Medical Research Center Jeddah, Saudi Arabia, ²Department of Pathology and Laboratory Medicine, King Abdulaziz Medical City, Ministry of National Guard Health Affairs Jeddah, Saudi Arabia, ³King Saud Bin Abdulaziz University for Health Sciences Jeddah, Saudi Arabia, ⁴Department of Laboratory Medicine, Faculty of Applied Medical Sciences Jeddah, Saudi Arabia

Acute leukemia has two major subtypes, acute myeloid leukemia and acute lymphoblastic leukemia (ALL). The central nervous system (CNS) is the most commonly affected extramedullary site in acute lymphoblastic leukemia (ALL). CNS involvement in ALL is classically defined by the presence of more than five leukocytes/μL and the presence of blasts on microscopic examination after centrifugation of the cerebrospinal fluid (CSF) sample. In routine practice, conventional cytomorphology (CC) is used for the detection of leukemic blasts on cytospin slides. This method is the gold standard for lymphoblast detection however it was estimated to have 96%–100% specificity but a low sensitivity for detection of CNS disease. CC could miss blast cells in patients with low count in the CSF, which need more sensitive method such as multicolor flow cytometry (MCF). Many studies demonstrated that MCF is a very important tool for detection of phenotypically abnormal cells and more sensitive than CC. The aim of this study is to compare the clinical utility of muticolor flowcytometry versus conventional cytomorphology regarding the early detection of leukemic cells in the CSF. Method Laboratory evaluation of the CSF samples obtained from the acute leukemia cases with low Leukocyte count in CSF at diagnosis by both Cytomorphologic examination and Multicolor flow cytometery Results This study included 308 acute leukemia cases, 31 were AML and 277 were diagnosed ALL cases. A 10.4% of the ALL cases were suspicious by CC, 37.9% out of them were proved positive by MCF. A 23.4% of the ALL cases that were negative for CNS infiltration by CC showed positive CNS involvement by MCF. As regards AML case 80% of the cases that were suspicious by morphology showed CNS involvement by MCF however 32.2%AML cases that were negative for blast cell infiltrate by CC were proved to be positive for CNS infiltrate by MCF. Conclusion: MCF showed better detection of CNS infiltration by leukemic cells compared to CC. more studies that include more acute leukemia patients and clinical correlation is recommended for further assessment of accuracy of CC compared to MCF and to assess the impact on patients outcome.

Corewell Health William Beaumont University Hospital Royal Oak, MI, USA

Introduction: Lupus anticoagulants (LA) are antibodies directed against phospholipids/phospholipid-protein complexes involved in coagulation and are associated with thrombotic events and recurrent fetal loss. There is no gold standard test for LA, therefore at least two LA-specific tests are recommended before excluding LA. Potential testing methods involve dilute Russell viper venom time (dRVVT) and hexagonal phase phospholipid assays (HPPA); however, many of these tests are susceptible to interference with anticoagulant therapy which can lead to false positive results. In this study, we compared the performance of our current HPPA, StaClot LA (SCLA), with a novel assay reported to have relatively minimal interference by anticoagulation therapy, CRYOcheck Hex LA (CCLA). Methods: We performed a retrospective study using 47 samples submitted for lupus anticoagulant testing in 3.2% sodium citrate vacutainer tubes. Samples were spun for 15 min at 3400 rpm to obtain platelet poor plasma. Samples positive by SCLA were selected for further testing and chart review for diagnosis of LA. These samples were frozen and retested by CCLA. We calculated the positive predictive value of both tests, assessed agreement between test methods using kappa statistics, and estimated the extent to which positive results were affected by anticoagulant therapy using Fisher's exact test. Statistical analysis was performed on R (version 4.2.1). Statistical significance was defined as p < 0.05. Results: Of the 47 samples, 34 were from women and 13 were from men, with an average age of 56 years. Using ISTH criteria, the presence or absence of LA was able to be determined in 42 out of the 47 samples. A 15 samples were positive for LA, and 27 samples were negative. Of the 15 positive samples, all were detected by both SCLA and CCLA. Of the 27 negative samples, all were positive by SCLA, and one was positive by CCLA. Thus, the positive predictive value for SCLA was 35.7% and 93.8% for CCLA. Since positive and negative results were obtained with CCLA, we were able to assess overall agreement between CCLA and true LA patients and CCLA and dRVVT, both of which were good (kappa = 0.78, p < 0.0001 for true LA; kappa = 0.73, p < 0.0001 for dRVVT). Additionally, 21 patients were taking anticoagulants at the time their sample was drawn. The rate of anticoagulation therapy was significantly higher in SCLA samples with false positive results (61.5% in false positive vs. 14.3% in true positive; p = 0.007). Conclusions: Diagnosis of LA can be challenging, particularly when the patient is being treated with anticoagulation. In our study, hexagonal phase phospholipid assays (SCLA and CCLA) showed high sensitivity (100%); however, SCLA showed a significant false positive rate, largely due to patients on anticoagulation therapy. CCLA on the other hand, shows promising resistance to anticoagulant therapy, and therefore, should be a strong consideration for labs offering LA testing in patient populations that are frequently anticoagulated.

¹Department of Clinical Pathology, Faculty of Medicine Hasanuddin University Makassar, Indonesia, ²Department of Physiology, Faculty of Medicine Hasanuddin University Makassar, Indonesia, ³Department of Internal Medicine, Faculty of Medicine Hasanuddin University Makassar, Indonesia

Sepsis is a life-threatening organ dysfunction caused by a disorganized host response to infection. Sequential Organ Failure Assessment (SOFA) is simple system, which can be used to identify organ dysfunction due to sepsis. Sepsis cause consumptive thrombocytopenia thereby stimulates thrombopoietin, which is a regulator of megakaryopoiesis dan thrombopoiesis to produce platelets. Immature Platelet Fraction (IPF) is one of the parameters to help diagnose sepsis in estimate platelets production dan differentiate thrombocytopenia caused by bone marrow failure due to toxic agents or systemic infection. This study was a prospective cohort study with the purpose of determining the levels of IPF and TPO at day 1, 3 and 5 with clinical severity of sepsis based on SOFA score. IPF levels were measured by flowcytometry method (Sysmex, Kobe, Japan) and TPO levels by ELISA method (Ray BioTech, Australia). There were 11 samples in this study and data were analysed statistically by repeated ANOVA test and Pearson correlation. The results showed that there was a significant positive correlation between increased levels of IPF and SOFA score with decreased platelet count (day-3, r = 0.753, p = 0.008) and no correlation between TPO levels and SOFA score in sepsis patients (day-1 p = 0.474, hari-3 p = 0.128, hari-5 p = 0.657). It was concluded that the higher IPF levels, the more severe clinical degree of sepsis.

¹Department of Biomedical and Molecular Sciences, School of Medicine, Queen's University Kingston, ON, Canada, ²School of Baccalaureate Nursing, St Lawrence College Kingston, ON, Canada, ³Department of Clinical Chemistry, Inselspital, Bern University Hospital, and University of Bern, Bern, Switzerland, ⁴Kingston Health Science Cancer Research Institute, Queen's University Kingston, ON, Canada, ⁵Clinical Pathology Department, Faculty of Medicine, Mansoura University, Mansoura, Egypt

Introduction: Thrombophilia is a disorder of blood coagulation that can lead to venous thromboembolism (VTE). While conventional coagulation assays may not be sufficient to assess VTE risk, global assays such as ROTEM may add important insights. Aim: We aimed to examine if ROTEM can detect hypercoagulability in patients referred for thrombophilia screening. Methods: A total of 376 patients were referred to a specialist outpatient unit for thrombophilia screening at an academic hospital. The average age was 43 ± 17 years with 38.6% (145) males and 61.4% (231) females. 66% (248) had at least one VTE, and 16.5 % (62) had recurrent VTEs. Additionally, 34% had a positive family history, 3.4% (13) had VTE in relation to pregnancy, and 28.6% were obese. In addition to routine thrombophilia testing, ROTEM was performed on blood samples and the coagulation patterns were evaluated using both extrinsic and intrinsic activators (EXTEM and INTEM). Hypercoagulability was identified as a shorter CFT, a larger AA and a greater MCF. Data compared between patients with and without previous VTE, using Mann-Whitney U and Chi-square tests and p < 0.05 was considered significant. Results: Patients who had a history of VTE demonstrated significant alterations in the coagulation parameters of EXTEM and INTEM (p < 0.001) compared to those without VTE (Tab.). Specifically, a statistically significant decrease in the clotting formation time (CFT), an increase in the alpha angle (AA), and an increase in the maximum clot firmness (MCF) were observed. 84 patients (22.4%) had known thrombophilia traits, while 145 (38.6%) had none. Conclusion: Our data suggest that ROTEM parameters, specifically CFT, AA, and MCF can detect hypercoagulopathy in patients referred for thrombophilia screening. Whether this translates into a higher risk for VTE recurrence must be determined in prospective cohort studies.

¹Laboratori Clinic Territorial ICS-IAS Girona, Spain, ²Group Neoma, Facultat de Medicina Girona, Spain, ³Hospital Meixoeiro, Vigo, Spain

Introduction: Neonatal lupus (NL) is an autoimmune disease characterized by the presence of antibodies anti Ro/SSA and La/SSB, which can across the placental barrier and cause lesions in neonates. It could appear as skin rash, heart block, liver dysfunction and citopenias. Most of the manifestations resolve by themselves, and the degree of cardiac involvement is the most important factor in mortality from NL. The diagnosis is mainly clinical and by the presence of specific antibodies in both maternal and fetal circulation. Case report 32 years-old pregnant diagnosed of Histiocytic Sweet's syndrome with positive anti-Ro, Anti-La and lupus anticoagulant with well-controlled pregnancy, negative serology. She gave birth to a full-term new born by normaldelivery to a mother. In his first hours of life, he begins with millimeter purpura maculae-type skin lesions on the face and more purplish and attenuated on the trunk, not palpable. From the emergency laboratory, thrombocytopenia was observed along with elevated liver enzymes. It was observed in the peripheral blood smear thrombocytopenia and the presence of small size mature lymphocytes with scanty cytoplasm, appearance of folded chromatin without apparent nucleolus. Twenty-four hours after birth, the patient presented general worsening with respiratory distress, hypertransaminasemia, and hyperbilirubinemia. Ventilation was carried out with CPAP enteric nutrition was started as well as broad-spectrum antibiotic therapy, with admission to the Intensive Care Unit (ICU) where treatment with intravenous gamma globulin was required as well as corticosteroids. Conclusion Neonatal lupus erythematosus is a rare condition that results from the passive transfer of autoantibodies from the mother during fetal life. The diagnosis of the disease represents a multidisciplinary challenge among pediatricians, laboratory specialists, cardiologists, rheumatologists, and dermatologists. In this case, the laboratory was of great importance both in the diagnosis and in the follow-up, evidencing an improvement at the hematological level (as observed with the increase in platelet levels), biochemical (e.g., in the values of liver enzymes) and Immunological (the titer of anti-SS-A and SS-B antibodies becoming undetectable).

¹Université catholique de Louvain, CHU UCL Namur, Haematology Laboratory Yvoir, Belgium, ²Université de Lorraine Nancy, France, ³Université de Namur, Department of Pharmacy, Namur Thrombosis and Hemostasis Center (NTHC), Namur Research Institute for Life Sciences (NARILIS) Namur, Belgium, ⁴Laboratoire mixte, CHU Dijon Bourgogne Dijon, France, ⁵Namur Thrombosis and Hemostasis Center (NTHC), Namur Research Institute for Life Sciences (NARILIS), Namur, Belgium

Background: Coagulolytic assays with (i.e., global fibrinolytic capacity of plasma (GFC)) or without the addition of exogenous fibrinolysis activator (i.e., euglobulin clot lysis time (ECLT)) exhibit distinct methodologies to enable clot lysis. On the one hand, the addition of exogenous t-PA overcomes the relative excess of PAI-1 in plasma but its sensitivity to hyperfibrinolysis remains to be evaluated. On the other hand, the removal of fibrinolysis inhibitors modifies the baseline equilibrium between endogenous PAI-1 and t-PA and might not reflect in vivo conditions, especially in the context of an inhibitor deficiency such as α2-antiplasmin. Few data are available concerning redistribution of fibrinolysis inhibitors and activators in the euglobulin fraction, and correlation between GFC and ECLT and fibrinolysis inhibitors and activators levels. Our aims were: (i) to measure the amount of several components of the fibrinolytic system in both platelet-depleted plasma and derived euglobulin fraction, (ii) to assess their correlations with ECLT and GFC. Methods: The study was approved by the local Ethics Committee (NUB: B0392020000042). Fibrinolysis components were measured in citrated platelet-depleted plasma from 20 healthy volunteers and derived euglobulin fraction, according to the experimental design summarized in Fig. 1a. Statistical analyses were performed in R (v4.1.2). Results: (1) A large reduction in α2-antiplasmin activity, α2-macroglobulin and TAFI antigen levels was observed in the euglobulin fraction as compared with platelet-depleted plasma (−87.7%, −100%, and −79.7% respectively), as expected (Fig. 1b). A massive reduction in PAI-1 activity level (−99.6%) was also noticed, despite a negligible reduction in PAI-1 antigen level (−9.2%) (Fig. 1b). (2) Moderate to very strong correlations were noticed between ECLT and plasma PAI-1 activity (ρ = 0.865, p < 0.0001) and antigen (ρ = 0.457, p < 0.05), and α2-antiplasmin activity (ρ = 0.462, p < 0.05) levels. Moderate to strong correlations between ECLT and t-PA antigen (ρ = 0.743, p < 0.001), and PAI-1 antigen (ρ = 0.559, p < 0.05) and activity (ρ = 0.477, p = 0.053) levels in the euglobulin fraction. (3) Moderate to strong correlations were observed between GFC and plasma t-PA antigen (ρ = 0.496, p < 0.05), α2-antiplasmin activity (ρ = 0.474, p < 0.05), and PAI-1 antigen (ρ = 0.669, p < 0.01) and activity (ρ = 0.690, p < 0.001) levels. Conclusion: Although the euglobulin fraction was mostly depleted in fibrinolysis inhibitors, PAI-1 was still present but in an inactive conformation, shifting the PAI-1/t-PA equilibrium and leading to the euglobulin clot lysis. Despite their distinct methodologies, ECLT and GFC were both mainly correlated with PAI-1 (antigen and activity), t-PA (antigen), and α2-antiplasmin (activity). Their sensitivity to fibrinolysis inhibitors deficiency remains to be studied.

¹CORE Laboratory. Biochemistry and Molecular Genetics Department, Biomedical Diagnostic Center Barcelona, Spain, ²Department of Mathematics. Technical University of Catalonia Barcelona, Spain

Introduction: Recognition systems based on deep learning are considered promising support tools in the diagnosis of hematological diseases. The rapid technological evolution in the field of deep learning leads to the adoption of new automatic recognition models. The objective of this work is to automatically differentiate eight types of cells in peripheral blood (PB) using the new Self-Attention approach of the Vision Transformer model. Methods: PB smears were stained with May Grünwald-Giemsa and cell images were acquired with CellaVision DM96. The data set from the images collected by our group and published in [1] included 3329 neutrophils, 3117 eosinophils, 1218 basophils, 1214 lymphocytes, 1420 monocytes, 2895 immature granulocytes (myelocytes, metamyelocytes and promyelocytes), 1551 nucleated red blood cells (NRBC) and 2348 platelets. We used data augmentation techniques to balance the number of images of each cell type. The database was divided into 65% for training, 15% for validation and 20% for testing. We designed a Self-Attention system based on Vision Transformer (Fig. 1). We used a model pretrained with the ImageNet-21K database and fine-tuned with the training and validation database. The model input was a set of images corresponding to PB cells (224 × 224 pixels). Each image was divided into 16 × 16 pixels patches and arranged in a corresponding position embedding. The embedding goes through the transformer encoder, the resulting vector is connected to a multilayer perceptron. The result was the classification of each individual image into one of eight groups. A total of 3419 images (test set) were used to analyse the performance of the model. Results: Using the test set, images of PB cells were automatically identified by the Vision Transformer model with the following accuracies: 97% (neutrophils), 100% (eosinophils), 99% (basophils), 99% (lymphocytes), 98% (monocytes), 97% (immature granulocytes), 99% (NRBC), and 100% (platelets). Fig. 2 shows the classification confusion matrix obtained using the test set. The overall accuracy of the system was 99%. Conclusion: Using a Self-Attention system based on Vision Transformer as it is described in this work, we obtained very high precision in the automatic classification of blood cell images. These results confirm the usefulness of applying new deep learning models as a support tool for peripheral blood cell recognition.

Albert Einstein Hospital São Paulo, Brazil

Introduction:: With the advancement of technology in recent decades, automation plays a fundamental role in clinical analysis laboratories. Exams that use automated systems present efficiency, reliability, reproducibility, and precision superior to manual methods. Although automated methodologies are already well established in the evaluation of peripheral blood, when it refers to bone marrow, there are several limitations that make it difficult to include some automated methodology for counting cells in bone marrow aspiration samples. This project aims to identify the limitations and possibilities of using the XN-10 automated equipment from the Sysmex® line in bone marrow cells counting. Methods:: 40 samples were selected through medical request for immunophenotyping and bone marrow aspiration exams. After the exams were performed, the remainder of filtered and diluted samples were analyzed in the Sysmex® XN-10 equipment. The results obtained were plotted, compared, and correlated with the results released in the requested exams. Results:: The results obtained from the analysis of bone marrow diluted in the Sysmex® XN-10 equipment were compared with the results obtained from the bone marrow aspirated, however, they did not show good correlation in any evaluated parameters: total mature granulocytes (R² = 0.2307), total erythrocytic series (R² = 0.2248), lymphocytic series (R² = 0.0892), monocytic series (R² = 0.2424), and total immature granulocytes (R² = 0.0859). Also, It was possible to compare some parameters with immunophenotyping and a poor correlation was obtained: mature granulocytes (R² = 0.3253), lymphocytic series (R² = 0.1461), and monocytic series (R² = 0.2817). Conclusions: The Sysmex® XN-10 equipment did not show good accuracy in the evaluation and differentiation of cells in bone marrow diluted samples.

University of Missouri, Department of Pathology and Anatomical Sciences Columbia, MO, USA

Introduction: The value of next generation sequencing (NGS) is frequently debated, yet few studies have shown the frequency where NGS can successfully identify genetic variants and available treatments. Sequencing a tumor by NGS may identify variants for which FDA-approved therapy exists for that tumor type, variants for which therapy is FDA-approved for a different tumor type, and available clinical trials, which exist for the reported variant. However, reporting from reference labs is not standardized, and not all pertinent information may be contained in the report. Methods: We reviewed NGS reports for all bone marrow specimens submitted to the University of Missouri between the years 2017 and 2022. Testing from Mayo Clinic (OncoHeme panel, 35 genes tested) occurred 2017 to 2021 (592 cases) and FoundationOne (CDx Heme, 406 genes (DNA) and 265 genes (RNA)) began in 2021 (91 cases). FoundationOne reports were reviewed for the presence of pathologic variants associated with an FDA-approved drug therapy for the patient's tumor type, a different tumor type, or available clinical trials. Mayo Clinic reports were reviewed for the presence of pathologic variants and their variant allele factor (VAF). Results: The FoundationOne reports reviewed demonstrated 48.4% of cases contained pathologic variants with some form of an available treatment. Within this group, 95.5% of cases had an available clinical trial, 27.3% had a treatment for their specific tumor type, and 52.3% had a treatment with benefit in a different tumor type. Of the Mayo Clinic reports reviewed, 75.2% of cases with an identifiable tumor type contained pathologic variants with a reported VAF. Failures of NGS due to insufficient tissue samples were identified in 17.6% of samples sent to FoundationOne and 5.2% of samples sent to Mayo Clinic. There are differences in reporting where FoundationOne reports do not include VAF in the report (available in a supplemental report), while Mayo reports fail to include clinical trials and possible drug treatments. Conclusion: NGS panels for hematologic tumors frequently identify pathogenic variants alone or with an associated treatment recommendation. These panels provide valuable results and the recent transition to FoundationOne gives additional information of available treatments and clinical trials. However, different reference labs provide different reports and there is a lack of industry standardization. This may impact the choice of laboratory selected for testing.

Ohio State

Mass cytometry enables quantification of up to 50 parameters per single cell through the use of time-of-flight mass spectrometry and heavy-metal tagged antibodies. This high number of parameters facilitates analysis of highly complex cell populations and allows for easy characterization of intracellular functional markers. There are, however, several unique challenges specific to mass cytometry that would require careful attention in any clinical assay. Fortunately, recent advances in mass cytometry methods have greatly improved the ease of generating highly accurate and reproducible data. Mass cytometry could be extremely useful for the diagnosis of malignant hematologic disorders, monitoring of minimal residual disease, and selection of treatment modalities. The capability of mass cytometry to easily measure cell cycle state and intracellular signaling would also make it ideal for performing functional immune monitoring, particularly for the assessment of organ or stem cell transplant and for the monitoring cellular or immunotherapy response. As with any high-parameter single-cell method, new data analysis approaches are required to take full advantage of richness of the data generated by mass cytometry assays. While challenges remain, mass cytometry could be an excellent tool for the functional assessment of clinical samples in high-consequence clinical decision-making assays.

¹Hospital of Meixoeiro Vigo, Spain, ²Siemens Healthcare S.L.U. Madrid, Spain, ³Siemens Healthcare Laboratory Diagnostics Surrey, United Kingdom, ⁴Siemens Healthcare Diagnostics Inc. Tarrytown, NY, USA

Introduction: The Atellica® HEMA 570 and Atellica HEMA 580 Analyzers (Siemens Healthcare Diagnostics Inc.) are newly launched quantitative, multiparameter, automated hematology analyzers for in-vitro diagnostic use in clinical laboratories. The classification and enumeration of blood cells is produced by impedance, double-hydrodynamic focusing, and flow cytometry-based white light absorbance and fluorescence detection technologies. The purpose of this study was to evaluate key hematology parameter recoveries between the Atellica HEMA 580 and Beckman Coulter DxH 800 automated hematology analyzers. Methods: Peripheral blood in K₂ EDTA from 175 samples with normal and various pathological results was analyzed on the Atellica HEMA 580 and Beckman Coulter DxH 800 systems within 4 hours of collection. The correlation of 14 complete blood count (CBC) and white blood cell differential (DIFF) parameters was evaluated by nonparametric (Passing-Bablok) regression. Results: Correlation coefficients ≥0.90 were obtained for white blood cells (WBC), red blood cells (RBC), hemoglobin (HGB), hematocrit (HCT), mean cell volume (MCV), mean cell hemoglobin (MCH), platelets (PLT), neutrophils (NEUT), lymphocytes (LYMPH), monocytes (MONO), and eosinophils (EOS). A correlation coefficient ≥0.80 was obtained for mean corpuscular hemoglobin concentration (MCHC) and ≥0.70 for red cell distribution width (RDW) and basophils (BASO). The slope for most of these parameters was very close to 1.0, indicating no statistical difference from concordance or identity. Conclusions: This preliminary study indicates close correlations of CBC and white blood cell differential parameters between the Atellica HEMA 580 and Beckman Coulter DxH 800 hematology analyzers. Atellica and all associated marks are trademarks of Siemens Healthcare Diagnostics Inc., or its affiliates. All other trademarks and brands are the property of their respective owners.

Hospital Israelita Albert Einstein São Paulo, Brazil

Introduction::Flow Cytometry (FC) is important for diagnosis and monitoring of chronic T- cell neoplasms (T-CLN). T cell neoplasms is challenging due to features with reactive T-cells and limitations of T-cell clonality tests. We validate, by FC, a 10 colors tube that includes TRBC-1 and compare with IOTest Beta Mark TCR Vβ Repertoire Kit (Kit). Methods: Fresh six peripheral blood and two bone marrow were received in EDTA with diagnostic hypothesis of T-CLN. These samples were stained with a 10 colors tube: Gamma/DeltaFITC/TRBC1(JOVI-1)PE/CD8ECD/CD3PC5.5/CD56PC7/CD2APC/CD7A700/ CD4A750/CD5PB/CD45KO. All samples were also stained with IOTest Beta Mark TCR Vβ Repertoire Kitwith CD3PC5.5/CD8ECD/CD4A750. Samples were acquired in Navios Flow Cytometer and analysed in Kaluza software (Beckman Coulter). Results: It was possible to detect clonality of T lymphocytes in seven samples and in four samples TRBC-1 was positive and in three samples TRBC-1 was negative. The Kit showed four samples with a single TCR-Vβ family and three samples with no expression of any TCR-Vβ family, suggesting clonality. In one sample T lymphocytes were considered polyclonal in both panels. In addition, two cases were positive for detection of clonal T cell receptor beta chain gene rearrangement by PCR. In the other cases, it was not possible to carry out the molecular study. Of the cases that presented clonality, three were T Cell Large Granular Lymphocyte Leukemia, two Sezary Syndrome, 1 Peripheral T Lymphoma NOS and 1 T Prolymphocytic Leukemia. Discussion: Detection of T clonality is not limited to molecular techniques and research laboratories that demand longer release times. The use of Kit presents good results and allows the classification of the TCRVβ family. However, it has a high cost, machine-specific adjustments, and analysis expertise. The JOVI-1 clone of the TRBC-1 antibody showed high specificity for the CRβ-chain constant region 1 (TRBC1) domain, providing a simplified immunophenotypic assessment of T cell clonality. Our 10-color tube enabled the detection of lymphocyte clonality T through the analysis of TRBC-1, by FC, proved to be a fast technique, easy to perform and with a short release period. Conclusion: The T clonality tube using the TRBC1 antigen constitutes a simple, inexpensive, and robust way to detect clonality of T lymphocytes by flow cytometry.

o

¹Ruxmaniben Deepchand Gardi Medical College Ujjain, India, ²Suraksha Diagnostic Center Kolkata, India

Introduction: Platelet counts generated by impedance methods (PLT-I) in cases of thrombocytopenia with interference flags can be unreliable. Alternative manual methods are used in such cases which has an inter–personal bias and requires expertise. Digital imaging of stained peripheral smears and then estimating counts on its basis minimizes these pitfalls of manual counting. We share our experience of evaluating utility of platelet count estimates in CellaVision DM1200 in cases of thrombocytopenia with platelet interference flag. Methods: Platelet count was done in nine fields (on CellaVision DM1200) with agreement of two observers on a digitized well stained smear of thirty peripheral blood samples having platelet count between 150 and 400(×10⁹/L). Conversion factor was calculated for each sample by dividing the count from the cell counter by the average PLTs/HPF value from CellaVison DM–1200. Derived conversion factor was 15.3. Further, consecutive 150 samples with platelet counts <120 × 10⁹/L were selected and were analysed in parallel using digitized smear technique (PLT-D) along with optical technique (PLT-O) in Horiba Yumizen H2500. Of these, 20 samples were also analysed using the international reference method (IRM). Statistical analysis using Pearson's correlation, student t-test, Bland-Altmann plot analysis and multivariate regression analysis was performed. Results: In the 150 cases evaluated, the mean platelet count (MPC) with PLT-I was 74.6 × 10⁹/L which was significantly lower than the MPC with PLT–D (114.2 × 10⁹/L). These counts correlated nicely with the MPC on PLT–O (108.9 × 10⁹/L). Pearson's correlation between PLT-O and PLT-D showed a good correlation (r = 0.809) and a significant correlation with IRM (r > 0.9). In Bland-Altmann analysis, PLT-D had lesser bias (8.225) and narrower limit of agreement (LOA) at 95% (−12.403 and 28.853) in comparison to PLT-I (Bias = 4.37; LOA = 2.902 to 84.498). According to CLIA proficiency testing criteria for acceptable analytical performance (±25% of target value), the platelet counts with PLT–D were compared with IRM and all the values were within 25%. The mean %difference between PLT-D versus IRM was 5.72%. Conclusion: Platelet count estimates using Digital platform (CellaVision DM1200) is rapid, simple, has better accuracy and minimizes inter-personal variation for reporting counts in cases of thrombocytopenia with platelet interference flags. This method is recommended as an alternative technique of manual platelet count on microscope.

Key Words: CellaVision; Digital; Platelet counts

¹Department of Medical Laboratory Science, College of Applied Medical Sciences Hail, Saudi Arabia, ²Department of Biology, College of Sciences, University of Hail Hail, Saudi Arabia, ³Department of Speech-Language Pathology and Audiology, College of Applied Medical Sciences, University of Hail Hail, Saudi Arabia, ⁴Department of Anatomy, Faculty of Medicine, King Abdulaziz University Jeddah, Saudi Arabia, ⁵Molecular Diagnostics and Personalized Therapeutics, University of Hail Hail, Saudi Arabia, ⁶Laboratory Department, King Fahad General Hospital, Jeddah Jeddah, Saudi Arabia, ⁷Pediatric Department, East Jeddah Hospital Jeddah, Saudi Arabia, ⁸Quality Department, King Khalid Hospital, Hail Hail, Saudi Arabia

Objective: Coronavirus disease 2019 is a communicable disease transmitted through the respiratory route. The severity of infection and mortality rate of COVID-19 cases was significantly high in the initial stages of the pandemic. This study aims to investigate the hematological profile of COVID-19 survivors and non-survivors. Method: In this single-center study, records of 108 patients hospitalized with laboratory-confirmed COVID-19 at East Jeddah Hospital between April-August 2020 were analyzed retrospectively for hematological parameters. Results: On comparison between survivor and non-survivor patients, the median length of stay in ICU was higher among non-survivors (7 days, IQR: 4.24 to 12) compared to survivors (5 days, IQR: 0 to 11.75) (p = 0.0151). The age positively correlated with the length of hospital stay (r(52) = 0.21, p = 0.0005) and ICU length of stay r(52) = 0.18, p = 0.001) for the survivor group. The median RBC counts were significantly higher in the survived group (4.56 × 10⁹/L, IQR: 4.02 to 5.11) in comparison with the non-survived (4.23 × 10⁹/L, IQR: 3.75 to 4.23) group (p = 0.0011). All COVID-19 patients exhibited lymphocytopenia and a significant negative correlation was observed between the lymphocyte values and length of hospital stay among the survived group (p < 0.001) as well as length of ICU stay among the survived group (p < 0.0480). Disease related mortality was significantly associated with reduced WBCs (8.5 × 10⁹/L, IQR: 6.1 to 11.7) and reduced basophils (0.09%, IQR: 0.02 to 0.19). A significant difference was also found between the survived and non-survived groups with respect to PT (12.5 sec. vs. 14 sec., p < 0.0001) and PTT (31.8 sec. vs. 40 sec., p = 0.0008). Conclusion: Hematological parameters can help to identify patients with severe COVID-19 and expected poor-prognosis/outcomes at the hospital admission stage. Cell counts of lymphocytes, WBCs, basophils and parameters such as PT and PTT can serve asclinical indicators of disease severity and progression to critical illness

¹Department of Pathology and Molecular Medicine, McMaster University Hamilton, ON, Canada, ²Department of Medicine, St. Joseph's Healthcare Hamilton, ON, Canada

Cleveland Clinic Cleveland, OH, USA

Introduction: Small B-cell lymphoproliferative disorders (B-LPD) may be difficult to subclassify in trephine biopsies, often requiring flow cytometry or excisional biopsy of an affected lymph node for definitive classification. Immunohistochemistry (IHC) can aid in classification, but not all B-LPD have specific positive markers. We sought to determine the utility of a panel of IHC for sub-classifying B-LPD. Methods: Cases each of hairy cell leukemia (HCL), chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL), follicular lymphoma (FL), mantle cell lymphoma (MCL), lymphoplasmacytic lymphoma (LPL), marginal zone lymphoma (MZL) and unclassified B-LPD were selected from the archives (07/19–03/22). IHC were performed on formalin-fixed, decalcified trephine biopsies:CD3 (2GV6, Ventana), CD20 (L26, Dako), MEF2b (EPR22193-78, Abcam), IRTA1 (EPR21961 Abcam), LMO2 (SP51, Cell Marque), CD10 (56C6, Cell Marque), BCL2 (124, Cell Marque), BCL6 (PG-G6p, Dako), BRAF V600E (VE1, Abcam), CD5 (SP19, Cell Marque), cyclin D1 (SP4, ThermoLabVision), LEF1 (EPR20294, Abcam), SOX11 (MRQ 58, Cell Marque) and CD200 (E519V, Cell Signaling Technologies). Each stain was blindly scored by two pathologists, with a third review for discrepancies. Results: 61 cases with definitive diagnoses, plus five unclassified B-LPD, were included. Among newer markers BRAF V600E was 100% specific (spc) for HCL, with 66.7% sensitivity (sen). For MCL, SOX11 was 100% spc and 70% sen. For CLL/SLL, LEF1 was 80% sen and 92.8% spc. IRTA1 was negative in all cases. Among germinal center markers, CD10 best identified FL (sen/spc 100%/92.8%) compared to LMO2 (40%/100%), MEF2b (90%/66.1%), and BCL6 (90%/85.7%). Used algorithmically, staining with 6 markers could positively identify 30 cases (6 HCL, 8 MCL, 8 CLL/SLL, 8 FL), and 22/22 MZL and LPL were negative for all algorithm stains, as expected (Fig. 1). Potential pitfalls in interpretation included LMO2 staining in hematopoietic cells and CD200 staining in follicular dendritic cells. Conclusions: IHC can correctly classify a significant number of cases, however many are still not classifiable with IHC, including MZL and LPL. The lack of specific markers for these entities remains a hindrance, but negativity for specific markers can indicate need for additional testing (e.g., MYD88 L265P). Optimization of some stains, especially newer markers (e.g., IRTA2, LMO2) for use in decalcified trephine biopsies may increase sensitivity.

¹Department of Pathology, Nova Scotia Health Authority (Central Zone) Halifax, NS, Canada, ²Department of Pathology, Dalhousie University Halifax, NS, Canada

Introduction: Morphologic assessment is the gold standard for the detection and quantification of blasts in peripheral blood, bone marrow, and body fluids. CellaVision is an automated digital microscopy platform routinely used for peripheral blood and body fluid morphology assessments, typically based on a 100-cell differential. Retrospective audits in our institution have identified blasts in blood films by manual microscopy after no blasts were detected using CellaVision. In these cases, the blasts were typically found outside of the morphology zone where CellaVision assessments usually occur. We hypothesized that increasing the number of white cells assessed, and therefore the size of the assessment area of the slide, would improve blast detection by CellaVision. Methods: Peripheral blood samples from healthy volunteers were inoculated with buffy coats from three acute leukemia patients. Blood films were made from serial dilutions of these manufactured specimens, and CellaVision differentials were determined from 100-, 200-, 300-, 400-, and 500-white blood cell captures. In addition, blood films found to have no blasts by 100-cell differential from 12 patients with a recent history of circulating blasts were subjected to repeat CellaVision assessments augmented by increased white blood cell capture. Results: The detection limit of blasts in the manufactured specimens was 0.26% with a 100-cell capture; this improved to 0.033% blasts when 500 white cells were assessed. Moreover, with a 300-white cell capture, blasts were detected in 10 of the 12 patient slides. The percent blasts detected in these patient specimens ranged from 0.3% to 3.8%; no blasts were detected in the remaining two patient slides, even when 500 white cells were interrogated. Conclusions: The ability to detect peripheral blood blasts using CellaVision depends on the number of white cells captured, which dictates the assessment area of the slide. This finding may reflect an unequal distribution of blasts throughout the blood film, whereby blasts are excluded from the usual assessment area of the slide when they are too sparse. Hospital laboratory services using automated digital platforms, such as CellaVision, to assess peripheral blood morphology should therefore consider increasing the minimum white blood cell capture to expand the blood film assessment area of peripheral blood films when screening for blasts.

Washington University

Spectral flow cytometry is a recently developed technology that allows for high-dimensional multiplexing of over 35 fluorophores in a single-tube assay. In this technology, fluorophores are excited by a series of lasers, and the emitted light is detected using several dozen sensors. Each fluorophore has a specific “light fingerprint” that can be unmixed, allowing simultaneous use of spectrally similar fluorophores, such as APC and AF647, that cannot be used at the same time with conventional flow cytometers. Functionally, this expands the menu of available fluorophores for panel design and makes it possible to use advanced machine learning analytic approaches to identify a wide variety of cell types and subsets. In this presentation, I will describe spectral flow cytometry technology, our experience using a 4-laser spectral flow cytometer with a 31-color panel in a research setting, and potential applications of this technology in the clinical flow cytometry laboratory.

¹Universitat Oberta de Catalunya Barcelona, Spain, ²Department of Mathematics. Technical University of Catalonia Barcelona, Spain, ³CORE Laboratory, Biochemistry and Molecular Genetics Department, Biomedical Diagnostic Center, Hospital Clinic Barcelona, Spain

Introduction: In recent years, the rise of artificial intelligence and its entry into the world of medical imaging has led to great advances. However, a common criticism of deep neural networks is that their way of storing knowledge is not easily exportable to forms more friendly to human understanding. For this reason, this work addresses the challenge of transforming knowledge of a semantic segmentation neural network into a set of numerical features easily understandable by humans, applied to normal and malignant blood cells. Methods A segmentation pipeline is developed using a total of 6000 digital cell images from peripheral blood (PB), acquired by the CellaVision DM96 and stained with May Grünwald-Giemsa. These images correspond to lymphocytes of healthy patients (N), reactive lymphocytes (RL), and abnormal lymphocytes corresponding to patients with follicular lymphoma (FL), chronic lymphocytic leukemia (CLL) and splenic marginal zone lymphoma (SMZL). The pipeline contains a training step of a deep neural network specialized in semantic segmentation called U-Net++ that has successfully separated the nucleus, the cytoplasm and an external zone around the cell. This model was training using the above data set and their corresponding ground truth masks generated by an automatic procedure described in [1]. A large sample of numerical features have been extracted from the entire dataset. As an example, both shape features (area, elongation, eccentricity…) and first-order statistical features of the pixel values (mean, standard deviation, kurtosis, among others) have been extracted. An app has been developed that, in a friendly manner, compares the current cell feature values (from a user's image) with the values of the database. Results: The semantic segmentation model has obtained a Dice coefficient (common metric to compare surface prediction efficiency) around 96% in just five iterations. Besides, an easy-to-use graphical interface has been built, which makes it possible to compare the features of the cell types at first glance. Fig. 1 shows a comparison of features between the current cell an abnormal LLC lymphocyte and normal lymphocytes population. Conclusions This work has developed a tool that allows the extraction of numerous features from the images so that the pathologist themselves can find relationships and patterns with morphological parameters, and thus understand more clearly aspects of the information with which they work.

References

[1] Alférez, Santiago, et al. "Color clustering segmentation framework for image analysis of malignant lymphoid cells in peripheral blood." Med. Biol Eng Comput 57 (2019):1265-1283.

Italian Society of Pathology and Laboratory Medicine–Hematology Working Group (SIPMeL-GdSE) Castelfranco Veneto, Italy

Background: The significant increase in Malaria disease (228 million cases according to 2018-WHO report), mostly due to globalization and migration from highly endemic countries (Africa, India), led to an increased incidence of these parasites observations, both in routine and emergency laboratory. Material and Methods: In October 2022 a survey on Malaria reporting was submitted to a selected and significant laboratories group by SIPMeL Hematology Group. 26 laboratories replied to the online survey, realized by SurveyMonkey® software, 20 days open, consisting of 38 single or multiple choice specific questions. The involved laboratories, distributed all over the country, were different for blood counts/year (range 1000 to 10 milions/year), automation and structure organization. Results: Main Survey items results are the following: parasites observed frequence: <1 case/month (87%); more cases found in 13%; diagnosis from a smear: 35% thin or thick smear; 31% both smears; 35% found in traditional blood smear; identification often suggested by instrumental data and flags, also generic ones; dedicated diagnostic guidelines (85%): 77% from the literature; 23% from local proposals; in 65% reporting rules shared among the team and 58% with the clinicians; in 21% specialist support need is required. outpatient specific requests: 69% for diagnostic suspicions; in 88% the case is handled as an energency; reporting traits: 54% morphology description, 31% semi-quantitative data; in 58% Parasites Count/GR ratio is provided; immunochromatographic rapid tests (88.5%): 50–100 parasites/μL blood sensitivity (72%;) falciparum specificity (47.6%); in 58.3 other plasmodia are distinguished (P. vivax: 15%; P. ovale: 12%; P. malariae: 12%); parasite DNA-PCR: 30.8% indoor performed; when not performed (69.2%) the reasons are cost (29.2%), low incidence (70.8%), specialized personnel need (20.8%); laboratory value definition in diagnosis: poor (12%); relevant (28%); essential (60%); Survey results agree with the laboratory test number, organization and equipment. Conclusions. Automation role is confirmed and considered as a fundamental contribution in screening and diagnostic investigation with abnormal data and flags for parasites presence/interference. Due to the magnitudo of the Malaria diffusion worldwide, the SIPMeL survey gives a good overview of the topic to dealt with, as a first step reaching widespread and common behaviors and as a base feature for improvement proposals and congruent intervention.

¹Department of Clinical Hematology, Hospital in HavÃÅ™ov HavÃÅ™ov, Czech Republic, ²Department of Hematooncology, University Hospital and Faculty of Medicine and Dentistry, Palacký University Olomouc Olomouc, Czech Republic

Introduction: Multiple myeloma (MM) is a plasma cell neoplasm characterized by clonal proliferation and accumulation of neoplastic cells and osteolytic skeletal involvement. Some of hemostasis disorders are attributed to M-Ig interactions with blood clotting factors (acquired von Willebrand's disease, acquired hemophilia A or deficits of other coagulation factors, circulating anticoagulant, hyperviscosity, amyloidosis and lupus anticoagulant) or with platelets (acquired thrombocytopathy), and also M-Ig-independent effects (thrombocytopenia, other thrombocytopathies, DIC, immobility and hypercalcaemia). The aim of our work is to detect abnormalities of coagulation in patients in with newly diagnosed MM suitable for intensit chemotherapy–depending on the aktivity of the disease, which predispose patients to thrombotic and bleeding complication, respectively, in MM. TGT is a global coagulation assai that measures the global capacity of blood plasma to form thrombin. Several clinical studies have shown that increased TG in platelet poor plasma (PPP) predicts an increased risk of (recurrent) VTE. Method We included 53 patients with newly diagnosed multiple myeloma in this study. Patients with MM were examined by coagulation tests for detecting both bleeding and thrombotic tendency with following coagulation tests: PT, APTT, TT, fibrinogen, antithrombin, D-dimers, levels of coagulation factors (II, V, VII, X, VIII, IX, XI and XII), vWF, lupus anticoagulant, protein C, protein S, resistance to activated protein C and trombin generation assai modified with activated protein C. We also monitored plasma cell counts and serum M-Ig levels in these patients. Results: All markers were evaluated (average value, standard deviation) to the dinase aktivity defined by the paraprotein level and a number of plasma cells (cytology analysis), respectively. A significant correlation was found between D-dimers and M-Ig quantity (p = 0.0031), D-dimers and plasma cells number (p = 0.0006), between vWF versus M-Ig quantity (p = 0.0053). No correlation was found between vWF and plasma cells number (p = 0.42), which is interesting. Correlations of vWF versus M-Ig quantity can predict bleeding conditions, however our ambition is to detect markers of thrombotic risk as well. For this purpose, we examined the modified TGT, which identified thrombotic pathology in eight cases (15%), while genetically determined thrombophilias were detected in only 3% of patients. Conclusion: In newly diagnosed patients with MM, were commend increased attention to the level of D-dimers and vWF, especially in patients with higher dinase aktivity according to M-Ig quantity in order to estimate possible bleeding or thrombotic complications and modified TGT for thrombotic complication, for which long term observation is needed.

Mayo Clinic

Inherited platelet disorders (IPDs) are a heterogeneous group of disorders characterized by normal or reduced platelet counts, bleeding diatheses of varying severities and presence (syndromic) or absence (non-syndromic) of involvement of other organs. Unfortunately, due to the lack of highly specific platelet functional tests, the identification of the underlying cause of IPDs remains clinically challenging and is often based on a clinical presentation and variable results of hematology and platelet laboratory testing results. Therefore, genetic testing has a potentially important role in the investigation of patients with IPDs. With the advent of next-generation sequencing (NGS) technologies, the rapid analysis of genes that have been previously implicated in IPDs is now possible. The potential limitations of NGS arise with the interpretation of the genetic information to definitively identify function-disrupting disease-causing variants. Thus, the interpretations of gene variants remain challenging, which will most likely be improved by thorough phenotype and genotype correlation studies and collaboration of the IPD clinical and research communities.

Hebei Yanda Lu Daopei Hospital Langfang, China

Background: Post-transplant lymphoproliferative disorders (PTLD) is a common and severe complication of allogenetic hematopoietic stem cell transplantation (allo-HSCT). Persistent EBV infection is one of main causes. Monitoring the abnormal proliferation of lymphocyte subsets by multiparameter flow cytometry (MFC) after HSCT may help to early diagnose PTLD. Besides, Flow cytometric sorting (FACSortig) may add the sensitivity. Methods we use MFC to detect immune subset immunophenotying, PTLD screening, and cytokines. Immune subset detection by two tubes panel of CD3/CD4/CD8/CD16+CD56/CD19/CD45 using truecount tube and CD45RA FITC/CD127 PE/CD4 PerCP/CD197 PECY7/CD25 APC/CD8 APC CY7/CD38 APC R700/CD279 BV421/CD3 BV510. PTLD detection by two tubes panel of ckappa FITC/clambda PE/CD38 PerCP/CD19 PE Cy7/CD10 APC/CD20 APC Cy7/CD56 bv421/CD45 V500 and kappa FITC/lambda PE/CD19 PE Cy7/CD5 APC/CD20 APC Cy7/CD45 V500. 24 kinds of cytokines were detected by CBA method. FACSorting panel is CD7 FITC/CD22 PE/CD3 PerCP/CD19 pecy7/CD138 APC/CD20 APC CY7/CD38 PB/CD45 V500c. One female patient of 64 years old who received allo-HSCT for severe aplastic anemia was continuously detected PB sample with EBV, MFC and FACSorting because she had low fever and cytopenia without lymph node enlargement at 45 days after HSCT. Result: the patient had low fever, cytopenia, and EBV positive by PCR at 45 days. Lymphosubsets immunophenotyping were detect by MFC on 30 days in which NK cells accounted for 90.11% and more than 90% of T lymphocytes were effector T cells. At 59 days B cells were 57/μL, NK cells and CD8+T cells increased simultaneously, but without increase of IFNγ and granzyme B cytomkines by CBA methods. Monoclonal cells plasma cells were present in PB at 70 and 77 days, and B cells were multiclonal(Fig. 1). The lymphosubsets in PB were sorted by FACSorting at 77 days, and isolated T, NK, B, and plasma cells were sent for further PCR detection. The results indicated that the copy number of EBV in B cells was 5.1 × 10⁶, none in NK, T and plasma cells. After antiviral therapy and gamma globule infusion, the symptoms and immune function recovered at 91 days, plasma cells were multicolnal and EBV were negative. The results at 117 days after transplantation indicated that the treatment was correct, and the abnormal proliferation of B cells and plasma cells was under control. (Fig. 1). Conclusion: The immune cell monitoring by MFC and FACSortingmay help to early diagnose EBV-related PTLD.

¹Hospital Ampang Ampang, Malaysia, ²Faculty of Medicine and Health Sciences, Universiti Sains Islam Malaysia Bandar Baru Nilai, Malaysia

The practise of prescribing Chronic Myeloid Leukaemia (CML) patients with indefinite tyrosine kinase inhibitors (TKIs) has gone uncontested, and the capacity of TKIs to eliminate the CML clone is also still unknown. Although the vast majority of CML patients do respond to TKIs, nonetheless, resistance may develop either de novo or during treatment. TKI resistance pathways are commonly classified as BCR-ABL1-dependent or BCR-ABL1-independent. The molecular evolution causal for this subset of individuals to lose molecular remission is still unknown. In this study, our goal was to explore the molecular mechanisms involved in resistance to TKI in patients who failed second-generation TKIs in order to identify potential genetic signatures and pathways that lead to TKI resistance. A preliminary of total 34 samples (two responder and 32 non-responder) were subjected to whole-transcriptomic analysis. mRNA gene expression, gene fusion and single nucleotide variation (SNV) were determined using Illumina DRAGEN Bio-IT Platform pipeline. Preliminary results for a total of 386 differentially regulated genes (DEGs) were identified, with 148 genes up-regulated and 238 genes down-regulated in the patients who failed second-generation TKIs (non-responder). The DEGs were enriched in histone acetylation/deacetlylation (HDACs/HATs), DNA methylation, and RNA polymerase regulation pathway. Further analyses demonstrated that CCCDC32-CBX3 fusion is significantly associated with non-responder. Techniques with enhanced sensitivity such as next-generation sequencing and the use of artificial intelligence techniques coupled with the development of mathematical modelling and computational prediction methods could reveal the underlying mechanism of drug resistance and facilitate the design of more effective treatment strategies for improving drug efficacy in CML patients. This preliminary analysis found that patients who failed second-generation TKIs express a potentially unique genetic signature. Analysis using larger sample size is necessary to validate these findings.

¹Hospital Ampang Ampang, Malaysia, ²Faculty of Medicine and Health Sciences, Universiti Sains Islam Malaysia Bandar Baru Nilai, Malaysia, ³Hospital Pulau Pinang Penang, Malaysia, ⁴Hospital Sultanah Aminah Johor Bahru, Malaysia, ⁵Hospital Tengku Ampuan Afzan Kuantan, Malaysia, ⁶Hospital Tuanku Jaafar Seremban, Malaysia, ⁷Hospital Melaka Melaka, Malaysia, ⁸Hospital Taiping Perak, Malaysia

The practice of indefinite tyrosine kinase inhibitor (TKI) provision for Chronic Myeloid Leukaemia (CML) has remained unchallenged. Furthermore, the ability of TKIs to eradicate the CML clone is still largely unknown. A multicentred observational study involving major hospitals in Malaysia to observe the clinical practise to End TKI In CML (EnTIC) was performed. The goal of this study is to determine the molecular response to TKI cessation in CML patients by close monitoring of BCR-ABL1.Adult CML in chronic phase patients who received first line TKI (Imatinib, Nilotinib or Dasatinib) for at least four years and achieved sustained MR4 (IS: 0.01%) for at least 2 years were recruited. A total 88 patients with equal gender distribution (44 male and 44 female) with a median age of 41 years old (range, 31.5–53.5 years old) were included in the study. These patients were observed with a median follow up of 14 months (range, 1–18 months). First-line TKIs were administered to all patients throughout the course of treatment, with median of 9 years (range: 7–13 years) and a median of 4 years for sustained deep molecular response (DMR) (range 3–6 years). After stopping TKI, 57 patients (or 64.8%) remain in treatment-free remission (TFR). Thirty-one patients failed TFR, with a median time of 2 months (range 1–5 months). Patients who remained in TFR at 166 days (5.5 months) and beyond were likely to achieve long term TFR in the following months. We observed the likelihood of loss of MMR during TFR trial is most often within the first 6 months of stopping TKI. However, a longer duration of follow-up and analyses of TFR data in EnTIC and other TFR studies will be needed to further evaluate the patients, and clinically significant characteristics before stopping treatment.

University of Utah and Intermountain Healthcare

It is important for clinicians who render neonatal care to precisely and reproducibly diagnose anemia. Confusion about the definition of neonatal anemia can arise because the diagnostic criteria are highly dependent on gestational and postnatal age. As an example, for a newborn infant delivered at 24 weeks gestation, a blood hemoglobin concentration defining anemia is much lower than a hemoglobin defining anemia at birth for a term neonate. Moreover, a hemoglobin concentration defining anemia at birth is higher than one defining anemia in that same infant 60 days after birth.

Diagnosing anemia can become evidence-based by using reference intervals derived from very large neonatal databases, comparing a patient's hemoglobin value with the appropriate gestational and postnatal age norms. To do this, we advocate defining anemia as a hemoglobin level that plots below the fifth percentile lower reference interval, defining moderately severe anemia as a hemoglobin value between the first and fifth percentile, and defining severe anemia as a hemoglobin plotting below the first percentile.

The information provided in this review can easily be adopted by clinical laboratories and by individual neonatal care units, thereby fostering application of these definitions for all infants who have a hemoglobin level drawn. Additional normative values included in this review that describe various other erythrocyte metrics can likewise be easily adopted. Doing this will enable a rigorous evidence-based diagnosis of neonatal anemia, and will also facilitate identifying the cause of the anemia, thus pointing the way to proper additional diagnostic testing and treatment.

Royal Devon University Healthcare NHS Foundation Trust, Exeter, UK

¹Royal Devon University Healthcare NHS Foundation Trust Exeter, United Kingdom, ²Global Marketing, CellaVision Lund, Sweden

Introduction – The Kleihauer–Betke (KB) method is commonly employed in transfusion laboratories for the estimation of fetomaternal hemorrhage (FMH) during pregnancy/post-delivery. It is a labor-intensive manual technique subject to high interobserver variability. While flow cytometry is considered the gold standard, it is often only available in specialist laboratories and often used only for confirmation of KB estimations. No automated device is currently available to analyze KB smears. The aim of this study was to develop a software for automated analysis of KB smears and evaluate its performance. Methods – A prototype software was developed on the CellaVision® DC-1 digital morphology analyzer. It analyzes tens of thousands of red blood cells (RBCs) and estimates FMH using artificial intelligence on smears prepared by the KB method. The software was developed and evaluated on smears using spiked blood samples with known concentrations of fetal RBCs. The precision was tested by comparing FMH estimates from the software before and after user review to flow cytometry on samples with 0.2, 0.5 and 1.6% of fetal RBCs. The effect of the analyzed number of RBCs on precision was tested using 10 slides from each concentrations analyzing 10 000, 25 000, 75 000 and 100 000 RBCs. A comparison study was performed on 30 samples, containing 0%–6% of fetal RBCs, and FMH estimates from the prototype software before and after user review were compared to that of flow cytometry. Results – The automated method using the prototype software is associated with acceptable within-run precision. The within-run precision of the method is augmented by increasing the analyzed number of RBCs, reaching an optimum at 50 000 cells. The automated FMH estimation using the prototype software produced a mean bias of 4.38 when comparing results before user review to flow cytometry. User review of the results on the prototype software significantly reduced the mean bias to 0.88 from 4.38. Conclusions – Our results show good correlation between the prototype software and flow cytometry, demonstrating the feasibility of using artificial intelligence for the estimation of FMH. Such an application could represent a viable and promising alternative to replace manual microscopy in clinical laboratories.

Laboratorio Analisi–S. Martino Hospital Belluno, Italy

INTRODUCTION The aim of this case report is to underline the need of integration between morphology and flow cytometry, even in cases apparently easy to diagnose. A blood sample of a 74 years old man, affected by small lymphocytic lymphoma (SLL) diagnosed 10 months before, was sent to our laboratory for flow cytometric immunophenotyping of lymphocyte subsets. We didn't know anything about his clinical conditions because he was admitted to another hospital. METHODS A complete blood count was performed on the analyzer Unicel DxH800 (Beckman-Coulter) to obtain lymphocyte count. The blood sample was analyzed on the flow cytometer Navios (Beckman Coulter) and tested for mature B cell neoplasm using: ClearLab 10C B Cell Tube (Kappa-FITC, Lambda-PE, CD10-ECD, CD5-PC5.5, CD200-PC7, CD34-APC, CD38-APC-A700, CD20-APC-A750, CD19-PB, CD45-Krome Orange) and a panel of appropriate antigens (CD11c-PC5.5, CD19-PC7, CD20-APC, CD22-FITC, CD23-PE, CD25-PE, CD43-FITC, CD45-KO, CD79b-PE, CD103-FITC) (Beckman Coulter). A blood smear was prepared for microscopic examination. RESULTS The lymphocyte count was 4.1 × 10⁹/L. B lymphocytes (1855 cells/μL) exhibited the classical SLL phenotype: CD5+, CD10−, CD11c−, CD20+ dim, CD22+ dim, CD23+, CD25−, CD38+, CD43+, CD79b−, CD103±, CD200+, with lambda light chain restriction. Conversely, the blood smear showed medium sized lymphocytes, with visible nucleolus, not suggestive of SLL (not clumped chromatin, no smudge cells). In the report for lymphocyte typing, we added morphologic evaluation of blood film in order to underline the evolution of SLL towards a large B cell lymphoma (Richter syndrome), despite the SLL phenotype. CONCLUSIONS We were asked to perform flow cytometry on a blood sample of a 74 old man affected by SLL, in order to confirm the diagnosis. The immunophenotype of B lymphocyte supported this diagnosis, but the evaluation of blood smear pointed out an early evolution to large B cell lymphoma. Later we found out the reason of patient's hospitalization: evening fever, cough, dyspnea, asthenia (symptoms suggestive of a more aggressive lymphoma than SLL). Subsequently, an inguinal lymphonode biopsy supported our assumption. Richter syndrome represents evolution of CLL/SLL to a large B cell lymphoma: the immunophenotypic profile of this disease can overlap with the one of CLL. So, even in presence of unequivocal flow cytometry results deposing for CLL/SLL, the review of the blood smear remains the starting point. Therefore, the answer to the title is that flow cytometry needs the integration with morphology, (even in non complex cases like SLL/CLL) in order to achieve a correct diagnosis.

Laboratorio Analisi–S. Martino Hospital Belluno, Italy

INTRODUCTION Green crystals of death are amorphous blue-green cytoplasmic inclusions found in neutrophils and monocytes in peripheral blood. This phenomenon occurs in critically ill patients, especially during severe hapatic disease. They are known as “crystals of death” since most patients quickly progress to death after their microscopic finding (92% within 72 h). We present the case of a woman (70 years old) who underwent orthopedic surgery for fracture of left femure. 2 days after osteosynthesis, she was transferred to Intensive Care Unit for unexpected hemorrhagic shock. Blood tests revealed sudden rise in aspartate aminotransferase (AST 5343 U/L), alanine aminotransferase (ALT 4101 U/L) and impairment in renal function (e-GFR 22.0 mL/min/1.73 mq). METHODS A complete blood count was performed (Unicel DxH800–Beckman-Coulter). A blood smear was prepared (DxH SMS Slidemaker–Beckman-Coulter) and then stained with May Grunwald-Giemsa staining (Aerospray 7152–Delcon SRL, Italy) for microscopic examination. RESULTS Blood count displayed anemia (Hb 85 g/L) and worsening thrombocytopenia (PLT decreased from 191 to 37 × 10⁹/L). The microscopic evaluation of blood smear was performed in order to investigate thrombocytopenia, which was confirmed. Blue-green, poorly defined, refractive, cytoplasmic inclusions were observed in rare neutrophils. CONCLUSIONS Green crystals of death are microscopic findings associated with acute hepatic failure, septic shock (E. Coli septicemia), multisystem organ failure secondary to trauma, severe SARS-COV infection. The pathogenesis of the blue-green inclusions is not well understood: current hypotheses suggest that they are made of lipofuscin. It is important to recognize these inclusions and differentiate them from others (e.g., Dohle bodies). However, even if they are thought to be predictors of imminent death, not all patients die within days of the crystals appearing: some patients recover and the crystal formation disappears from blood sample. In this case report, the patient was diagnosed as acute hepatic failure secondary to hypoxic injury. Green crystals of death appeared when AST and ALT reached the maximum values (5343 U/L and 4101 U/L, respectively) and then disappeared from blood as soon as AST and ALT decreased. Later, her clinical condition improved: she was discharged from hospital and is actually doing physical therapy. Therefore, these blue-green inclusions correlate with the severity of clinical status but are not always associated with the patient's death.

Catholic University of Sacred Hearth, Rome, Italy

The date of birth of modern hematology laboratories can be traced back to 1953, when Wallace H. Coulter patented the first system capable of counting red and white blood cells automatically, also calculating Wintrobe's three erythrocyte indices. The Coulter principle (or effect) was based on the poor ability of blood cells immersed in a flow of saline solution to conduct an electric current through an orifice (impedance), thus blocking its passage for a time-dependent on their volume. In the 1960s and 1970s, instruments based on optical detection principles were introduced, such as all light scattering and absorbance, which also perfected platelet counting. Even more recently, dyes, often fluorescent, capable of distinguishing and counting reticulocytes, were used. The automatic differential count of leukocytes obtained thanks to heterogeneous methods such as automated cytochemistry, measurement of scattering at different angles, and fluorescence now has excellent reproducibility and accuracy for the five classes of normal circulating leukocytes, sensitivity and specificity appropriate for the screening of pathological samples to be reviewed under a microscope and, in the latest generation equipment, also the ability to count, with variable accuracy, some populations of pathological cells, such as immature granulocytes (promyelocytes, myelocytes, metamyelocytes), blast cells and erythroblasts circulating, thanks to the specific conformation of the nuclei and cytoplasm of these elements. New erythrocyte indices have been developed for the differential diagnosis of anemias, such as the percentage of hypochromic, hyperchromic, microcytic and macrocytic red blood cells, the mean hemoglobin content of red blood cells (CHCM) and reticulocytes (CHr, Ret-HB and similar). Different subpopulations of cells with different nucleic acid content (immature fraction of reticulocytes or IRF, cross-linked platelets) are usually inversely proportional to the life span and rate of production of those cells. Each instrument now provides numerous additional parameters that numerically translate the position of cells on the cytogram space and measure their size, granularity, and other physical or chemical properties. Numerous researchers are working on trying to apply such population data to particular cell types or hematological diseases. An interesting result was obtained using the constant neutrophil/lymphocyte ratio in patients with COVID-19, as this ratio clearly highlights the more severe cases, simultaneously characterized by lymphopenia and neutrophilia. Hematology analyzers can currently be used to count cells in nonblood body fluids and even, with some limitations, in bone marrow fluid. We are also recently seeing a revival of methods based on digital morphological analysis of peripheral blood cells and automated microscopy. The artificial intelligence programs used for this purpose are still relatively rigid and, in any case, linked to human intervention. Telemicroscopy, morphological education and the second level of routine microscopy greatly benefit from this morphology recovery with digital techniques.

Reference. Automated hematology analyzers: State of the art. Brugnara C and Kratz A, Eds, Clin Lab Med 2015;35(1):1-224.

Department of Clinical Laboratory, Children’s Hospital of Nanjing Medical University, Nanjing, China

Background: Serum amyloid A (SAA) is a class of lipid-binding protein in plasma with great value in indicating possible infectious diseases, cardiovascular diseases, and prognosis of tumor patients. The hook effect impairs laboratory detection of high values of SAA. Mindray has launched the new BC-7500CS automated hematology analyzer with an integrated SAA autodilution (SAA-D) function to reduce the hook effect. This study aimed to verify the performance of the SAA module, especially in the high SAA samples. Methods: Blood samples were randomly collected from outpatients and inpatients of the Children's Hospital of Nanjing Medical University (CH) and analyzed with the SAA module with the BC-7500CS automated hematology analyzer. Two hundred twelve blood samples with SAA concentrations higher than 100 mg/L were collected to verify the SAA-D function by comparing with the manual dilutions. The reportable range of the instrument was investigated by measuring the samples with different concentrations of SAA (up to 2000 mg/L). Interference studies were conducted with blood samples with added bilirubin (BIL), triglyceride (TG), vitamin C (VitC), rheumatoid factor (RF), human anti-mouse Antibody (HAMA), hemoglobin (HGB). The basic performance (e.g., background, repeatability, etc.) of the SAA mode was also validated. Results: The basic performance of the SAA modules of Mindray BC-7500 CS all met the manufacturer's claims. The maximum measurable SAA was 1932.38 mg/L, and the SAA module showed high anti-interference ability when BIL <100 mg/dL, TG <1408 mg/dL, VitC <150 mg/dL, RF <65 IU/mL, HAMA <1000 IU/mL, and HGB <203.1 g/L. In addition, the SAA-D showed a good correlation with the manual dilution for samples with high SAA concentrations (Y = −26.03 + 1.041X, n = 212, r = 0.993). Conclusions: The Mindray BC-7500CS showed good performance in measuring SAA concentrations, and the SAA-D function could accurately measure high SAA concentrations, enabling reliable SAA detection in the laboratory and clinical practice.

¹Department of Biology, Faculty of Science, Mahidol University Bangkok Thailand, ²Biomedical Science Program, Faculty of Science, Mahidol University Bangkok Thailand, ³Department of Biology, Faculty of Science, Khon Kaen University Khon Kaen Thailand, ⁴Division of Bioinformatics and Computational Biology, Knight Cancer Institute, Oregon Health and Science University Portland, OR, USA, ⁵Department of Cell, Developmental and Cancer Biology, Knight Cancer Institute, Oregon Health and Science University, Department of Cell, Developmental and Cancer Biology, Knight Cancer Institute, Oregon Health and Science University Portland, OR, USA

Introduction Acute myeloid leukemia (AML) is a fast-progressing hematopoietic malignancy and the most common leukemia in adults. Hyperactivating mutations of fms-like tyrosine kinase 3 (FLT3) are the most common genetic alterations found in AML and are related with poor prognosis. Development of FLT3 inhibitors (FLT3i) have improved treatment outcomes in AML patients. Unfortunately, the benefit is short-term due to acquired resistance. This urges the need to find novel treatment approaches. Recently, there have been several studies demonstrating that combining FLT3i with other small molecule inhibitors including MEK inhibitor (MEKi), enhanced anti-leukemic activity in AML cells. Nevertheless, whether different AML subgroups with distinct genetic backgrounds will response to this combination drugs and whether resistance can develop are not known. Methods To identify genes involved in response to combination of FLT3i and MEKi, genome-wide CRISPR knock out (KO) screen was performed on FLT3 mutation harboring MOLM13 AML cell line. Results CRISPR screen identified PTPN1, a negative regulator of JAK/STAT signaling pathway, as one of top hits that loss of function promoted AML cell survival in the present of the combination drugs. This suggested hyper-activation of JAK/STAT signaling as a potential indicator predicting resistance toward this combination strategy. PTPN1 KO cells were generated using single gRNAs to examine role of PTPN1 in response to FLT3i and MEKi. Drug sensitivity assay demonstrated that PTPN1 KO AML cells were resistant toward FLT3i, gilteritinib, and MEKi, trametinib. Although, ex vivo functional analysis of 163 AML patient samples did not showed correlation between RNA expression of PTPN1 and response to this combination drugs, higher expression of JAK2 but not JAK1 and JAK3 was correlated with resistance toward combination of FLT3i and MEKi. Conclusions Taken together, our results proposed that JAK/STAT signaling can act as a mediator of resistance toward combination of FLT3i and MEKi and that AML patients with hyper-activation of JAK/STAT signaling will not be benefit from this combination strategy.

National Hospital Kandy Sri Lanka Kandy, Sri Lanka

Introduction: Plasma cell neoplasms as Multiple Myeloma are malignancies derived from mature B cells. Aberrant T cell antigen expression is extremely rare in those and only a few such cases have been described. We describe a case of Multiple Myeloma with aberrant CD3 expression, which was detected by flow cytometry. Case Report A 77 years old man presented with back ache for two months and found to have ESR of 130 mm/first hour and haemoglobin of 74 g/L and serum creatinine of 300 Âμmol/L. His serum protein electrophoresis revealed a monoclonal band of 20 g/dL in É£ region and immunofixation showed Immunoglobulin G Kappa (IgGƘ) monoclonal band. Bone marrow biopsy and flow cytometry were performed, and marrow morphology showed >90% plasma cells, out of which 17% were plasma blasts. Flow cytometry demonstrated Kappa (Ƙ) restricted cell population with CD38, CD138 strong positivity and CD20dim positivity. CD 19, CD 10 and CD 79a were negative in this population. CD 3 was aberrantly positive in these cells and no other T cell markers were detected. Patient was treated with Bortezomib, Thalidomide, low dose Dexamethazone with GCSF support and after four doses of Bortezomib his serum creatinine markedly improved and serum protein electrophoresis became normal with no detectable monoclonal band. Response assessment can also be done by quantification of abnormal plasma cells using aberrantly expressed immunophenotypic markers. Discussion Although plasma cell neoplasms are known to express antigens of different lineages, expression of T cell related antigens as CD3, CD4 are extremely rare. Pathobiology of expression of T cell antigens in plasma cell neoplasms are described with multiple theories. Some of these are EBV infection, lineage infidelity and PAX 5 down regulation. It is also described that T cell antigen may express in recurrence of Multiple Myeloma. Furthermore, it is identified that response assessment of Multiple Myeloma can be done using monitoring of paraprotein level as well as quantification of abnormal plasma cells using aberrantly expressed immunophenotypic markers. This case emphasizes the importance of recognizing rare immunophenotypic aberrancies in plasma cell neoplasms as Multiple Myeloma, which may cause diagnostic difficulties. Key words: CD3 positive Myeloma, plasma cell neoplasms

¹SAMRC/CPUT Cardiometabolic Health research unit, Cape Peninsula University of Technology Cape Town, South Africa, ²Division of Haematology, University of Cape Town, Cape Town, South Africa

Introduction: Sepsis is characterized by multi-organ failure due to uncontrolled immune responses to infection. Prevalence is increased in developing countries and requires prompt diagnosis and treatment. Clinical symptoms are often nonspecific, with blood cultures and procalcitonin investigations being adopted as diagnostic tools in clinical practice. In resource limited settings these tests are often unavailable and therefore the need for an efficient, cost-effective method to screen for early sepsis is important. Reports, although controversial, especially in Africa where benign ethnic neutropenia is common, suggest that the calculation of cell ratios could assist in the early identification of sepsis. This study aimed to evaluate the use of the neutrophil to lymphocyte (NLR), monocyte to lymphocyte (MLR) and platelet to lymphocyte (PLR) ratios as a screening tool for sepsis in hospitalized patients. Methods: Complete blood counts, blood cultures and biochemical tests of 125 consecutive adult patients admitted to a hospital in Cape Town were retrospectively analyzed. All tests were performed in an ISO15189 accredited laboratory and those with a positive blood culture and procalcitonin (>2 ng/L) were considered as having sepsis. Utilizing the automated differential counts, the NLR, MLR and PLR ratios were calculated. The results were compared between those with and without sepsis and correlated with procalcitonin levels. A p-value of <0.05 was considered significant. Receiver operating characteristic (ROC) curves were constructed to evaluate the cell ratios as possible predictive markers for sepsis. Results: Sixty two patients with sepsis were identified and compared to sixty three non-sepsis controls. All cell ratios were significantly elevated in sepsis patients (p < 0.001), however, no significant difference in the absolute monocyte counts (p = 0.377) was detected. No correlation between any of the cell ratios and procalcitonin was observed, while ROC curve analysis established the NLR as having the best predictive value (AUC: 0.977; Youden index: 0.855). The MLR was considered as a good marker of sepsis (AUC: 0.877; Youden index: 0.710), while the PLR, only performed moderately (AUC: 0.699; Youden index: 0.455). Conclusion: The full blood count is widely available, inexpensive, and routinely requested by emergency care clinicians. Although procalcitonin and blood culture remain the gold standard, the calculation of cell ratios, when used in combination with other full blood count parameters, could provide a simple screening tool for the early detection of sepsis particularly in resource limited settings.

Hospital de ClÃnicas da Universidade Federal do Paraná Curitiba, Brazil

Introduction: Sensitivity of Minimal/Mensurable Residual Disease (MRD) methods is increasing in last years. There are few studies detailing the bone marrow microenvironment and kinetic of B-cell reconstitution in the setting of hematopoietic stem cell transplantation (HSCT). Objective: This prospective study aims to analyze multiparameter flow cytometry (MFC) sensitivity follow by analysis of medullary B-cell maturation kinetic after transplant in both ALL and AML patients. Methods: prospective analysis 64 consecutive patients transplanted between 2019/January and 2022/June, with MRD evaluation in the month before HSCT, d30, d60, d90 and d360. MFC was using a 10-color tube included the backbone markers CD19, CD45, CD34, CD10, CD20, CD38 and CD81 and four discriminatory markers (CD66c and CD123 in PE, CD73 in BV605, and CD304 in APC-R700). FACSLyric™flow cytometry's and Infinicyt® were used to MFC analysis. The achieved sensitivity and percentages of normal mature and precursor B-cells, plasmatic cells, myeloid precursors, and stromal cells were compared before and after transplantation. Statistical analyses were performed in Graph Prism v7.0. Results: 34 ALL (42.6%) and 41 AML (57.4%) were transplanted in this period. Among 64 patients in morphologic remission, we found 38 (59.4%) MRD negative and 26 (40.6%) MRD positive before transplant. Posttransplant MRD analysis was possible in 26 patients on d30, 24 on d60, 57 on d100 and 18 on d360, and maturation analysis in 84 samples of 41 patients (26 BCP-ALL, 7 high risk T-cell leukemia, and 8 AML). Sensitivity of ≤0.01% was achieved in 81 of 84 samples (96%). All but two samples had more than 9 000 000 cellular events acquired, the mean LoD 0.004% (range 0.0005%–0.028%) and LoQ 0.01% (range 0.001%–0.07%). All patients had the same percentages of lymphoblasts, normal mature and precursor B cells, plasma cells in the pre- and post-transplantation assays. No differences in post-transplant B-cell recovery between days d60, d100 or d360 post-transplant. Mature and B-cell precursors were significantly low on d30 compared to the other days, confirming a temporal maturation of these cells starting on d60 in all leukemia types. Conclusion: This study showed this strategy achieves maximum sensitivities at various time points after bone marrow transplantation. Furthermore, the kinetics of lymphoid, plasmatic, myeloid, and stromal cell maturation were similar across leukemia types and transplant time points in this cohort. There was a decrease in B-cells immediately post-transplant in all groups, with progressive recovery from d60 onwards, reflecting the immune recovery typically seen in this setting.

Emory University, USA

Conventional cytomorphologic assessment is still essential among growing numbers of technologically more sophisticated tools available to the practitioners of the discipline of hematopathology. It is recommended for visual numeration of blasts, when required, and widely accepted as a cost-effective approach to triage additional ancillary testing. Of interest, certain cytomorphologic features have been found to be associated with varied hematologic diseases and therefore alert to investigate such during the diagnostic work up of blood and bone marrow samples. Of these, the presence of vacuoles in certain hematopoietic cell types have been described in copper deficiency and the newly reported autoinflammatory disease with acquired somatic genetic mutation VEXAS (vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic) syndrome, among others. The diagnostic implication, disease association, and differential diagnosis of the presence of vacuoles in certain hematologic diseases are being discussed and highlighted by illustrative clinical case studies.

Ghent University Hospital

Diagnosis of Acquired Thrombotic Thrombocytopenic Purpura

Thrombotic Thrombocytopenic Purpura (TTP) is a rare and life-threatening thrombotic microangiopathy characterized by microangiopathic hemolytic anemia, severe thrombocytopenia, and organ ischemia linked to disseminated microthrombi. TTP is specifically related to a severe deficiency in ADAMTS13, the von Willebrand factor-cleaving protease. ADAMTS13 deficiency is most frequently acquired via ADAMTS13 autoantibodies, but rarely, it is inherited via mutations of the ADAMTS13 gene. The idiopathic autoimmune form of the disease is the most common.

The first acute episode of TTP usually occurs during adulthood, but TTP begins in childhood in the inherited forms. Rapid recognition of TTP is crucial to initiate appropriate treatment. Diagnosis relies on the combination of the clinical symptoms, the disturbed laboratory parameters and the measurement of ADAMTS13. Prompt identification/exclusion of TTP can be facilitated by rapid ADAMTS13 testing. Diagnostic assays are available to identify ADAMTS 13 activity. The most commonly utilized enzyme-linked immunosorbent (ELISA) assay takes several hours to perform and so does not generally permit rapid testing. Nowadays, automated tests are available to measure ADAMTS13 activity. Also, acquired antibodies to the enzyme can be measured. Antibodies to ADAMTS 13 can be demonstrated in about 75% of cases of acquired TTP associated with ADAMTS 13 activity levels <10%. These antibodies reduce circulating functional enzyme levels. Most autoantibodies are inhibitory and therefore can be detected and titrated in vitro using classical mixing studies. The use of recombinant ADAMTS 13 in an ELISA allows the rapid identification of autoantibodies, primarily IgG.

A long-term follow-up of patients with TTP is crucial to identify the occurrence of other autoimmune diseases, to control relapses, and to evaluate clinical sequelae. Long-term follow-up of patients with TTP, includes medical consultation, standard laboratory tests, and ADAMTS13 activity monitoring.

Washington University, USA

Genomic analysis is essential for risk stratification in patients with acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS). Whole genome sequencing (WGS) has the potential to greatly simplify and improve conventional multi-modality diagnostic approaches involving cytogenetics, FISH, and panel-based sequencing by allowing for comprehensive genomic evaluation in a single rapid assay. While WGS has historically been too complex and too expensive for use in the clinical laboratory, recent advances in sequencing and data analysis coupled with rapidly decreasing sequencing prices have made WGS an attractive alternative to conventional testing. This presentation will focus on key technical innovations that have allowed the routine use of WGS in the clinical laboratory as well as data demonstrating the clinical utility of WGS compared to conventional testing.

Washington University

Measurable (or minimal) residual disease (MRD) has become an important biomarker in the management of acute myeloid leukemia (AML) and other myeloid malignancies. While AML MRD has been historically measured by flow cytometry, interpretation of data is complicated and is limited to cases with blast phenotypes that can be distinguished from normal. Early molecular MRD methods using PCR-based identification of recurrent translocation (i.e., PML::RARA) have shown a high sensitivity and high predictive value for identifying relapse, however, these methods are limited to the subset of AML patients with recurrent translocations. Newer sequencing-based MRD assays which target gene level mutations present in >90% of AML cases have had a major impact in the management of AML. Recent advances in error corrected sequencing and drops in sequencing costs have made sequencing-based MRD assays accessible to more laboratories. This talk will focus on methods for sequencing-based MRD, clinical interpretation of MRD data, and how MRD data is being used to guide treatment decisions.

Washington University, USA

ISLH is a unique organization, promoting education in laboratory hematology by collaborating with colleagues and societies throughout the world. This presentation will review the evolution of ISLH from its inception 35 years ago to the present, highlighting significant events and innovations that sustain ISLH's mission. Laboratory Stewardship is an ideal that is widely embraced while challenging to implement. The second part of this presentation will address some of the underlying principles and values of laboratory stewardship from the perspective of one academic medical center in the United States.

The University of Western Australia

The myeloproliferative neoplasms (MPN) are a group of clonal stem cell disorders with similarities at the phenotypic and molecular level. Clinically, they are characterised by over-production of one or more mature myeloid elements and a variable tendency to develop acute myeloid leukaemia (AML). Polycythaemia vera (PV), essential thrombocythaemia (ET) and primary myelofibrosis (PMF) overlap clinically and share a tendency to undergo phenotypic shift, such that patients with ET may develop PV, and ET or PV may undergo myelofibrotic transformation. In this presentation the morphology of PV, ET and PMF will be described. The presentation will focus on the blood count and film appearances as well as bone marrow cellularity, pattern, megakaryocyte morphology and reticulin content.

¹Department of Haematology, Tan Tock Seng Hospital Singapore, Singapore, ²Department of Laboratory Medicine, Khoo Teck Puat Hospital Singapore, Singapore, ³ASUS Global Pte Ltd Singapore, Singapore, ⁴Department of Laboratory Medicine, Tan Tock Seng Hospital Singapore, Singapore, ⁵ASUSTeck Computer Inc. Taipei, Taiwan

Introduction: Manual light microscopy (MM) of peripheral blood film (PBF) is tedious and time-consuming. A deep-learning Artificial Intelligence (AI) model, Blade, has been utilized to identify white blood cells (WBC) through a convolutional neural network (CNN) trained by semi-supervised deep learning technique using 185 412 peripheral blood film cells. We compared the performance of Blade with a commercial model, the Cellavision DM9600 (Cellavision) in a real world laboratory setting. Methods: 168 films were randomly selected from the routine laboratory bench to obtain a WBC differential count by evaluating 200 WBC per film, using MM, Cellavision and Blade. 6 lab-certified medical technologists reviewed these films on MM, followed by assessment by Cellavision, and then by Blade. The study assessed 13 categories: Neutrophil, Lymphocyte, Large Granular Lymphocyte, Reactive Lymphocyte, Monocyte, Eosinophil, Basophil, Metamyelocyte, Myelocyte, Blast, Smudge Cell, Artifact, Giant Platelet. Correlation (r, r²) of differential count was plotted for both test methods against MM which was considered as the gold standard. Agreement between AI test methods and the reference method (MM) is calculated for various abnormalities (Fig. 1). The performance (accuracy, sensitivity, specificity, and precision) of manual microscopy and AI based test methods (MM, Cellavision, Blade) for the detection of abnormal films was evaluated. Results: Blade's correlation to MM for correct identification of common cell-types is higher than that of Cellavision, ranging between 0.72 to 0.98 for Blade, and 0.62 to 0.97 for Cellavision (Fig. 1). Agreement between test method and MM for characteristic differential count findings (abnormalities) show Blade with a higher agreement in detecting Granulocytosis, Lymphocytosis, Lymphopenia, Myelocytes, Metamyelocytes and Blasts. Conversely, the agreement is higher in Cellavision for Eosinophilia and Monocytosis (Fig. 1). The performance of Blade in detecting abnormalities is 87.5%, 92.3%, 70.3%, 91.7% (accuracy, sensitivity, specificity, and precision). This is higher than that of Cellavision's DM9600, being 83.9%, 90.8%, 59.5%, 88.8%, respectively. Statistical assessment proved non-inferiority of Blade to Cellavision, given a two-sided level of significance of 5%, a 70% power and a 5% limit. One possible confounding factor is that different regions of the same film are analyzed on test and reference methods, which can lead to variation. Conclusion: The performance of Blade is higher than that of Cellavision. It is non-inferior and comparable in overall performance to Cellavision DM9600. Local laboratory data can possibly be used to augment existing algorithms and train existing commercially available solutions for a higher accuracy in WBC identification.

¹Biochemical and Molecular Genetics Department, CORE Laboratory, CDB, Hospital ClÃnic of Barcelona, Barcelona, Spain, ²Clinical Institute of Internal Medicine and Dermatology, ICMIiD, Hospital ClÃnic of Barcelona, Barcelona, Spain

Introduction: Glucose-6-phosphate dehydrogenase (G6PD) deficiency, the most prevalent enzymopathy in the world, is an X-linked genetic disorder that affects the pentose phosphate pathway. There is wide inter- and intra-individual variability in the development of haemolytic crisis and several factors influence it. A patient with an acute haemolytic episode and methemoglobinemia caused by eating broad beans for the first time in adulthood is presented. Methods: A 75-year old female patient was admitted to the emergency room with a 2-day history of progressive jaundice, dark urine, anorexia and generalized arthromyalgia. She did not refer fever or abdominal pain, but she suffered of an oppressive nocturnal thoracic pain and progressive dyspnea. Blood cell counts and hematological parameters were obtained using the Advia®2120i analyzer. Peripheral blood (PB) smears were stained with MGG and blood cell images were acquired using the CellaVision DM96. The G6PD activity in erythrocytes was measured. Confirmation of carrier status was done by genetic study using PCR Sanger sequencing. Results: Laboratory data revealed high WBC count (16.7 × 10⁹/L), low haemoglobin values (95 g/L) and normal platelet count. High reticulocyte values (548.4 × 10⁹/L) were found. Blood gases analysis revealed SatO₂ of 97.1% and mildly increased levels of methemoglobinemia (5.2%). An increased C-reactive protein (10 mg/dL), hyperbilirubinemia (7.5 mg/dL and unconjugated bilirubin of 7.1 mg/dL) and high troponin values were observed. On suspicion of haemolytic anaemia, a Coombs test was requested, which was negative. The PB smear review showed macrocytosis, polychromasia, spherocytes and excentrocytes (Fig. 1). After all these findings, a G6PD deficiency was suspected. The patient was re-interviewed and she explained consumption of fava beans during the five previous days. G6PD enzyme activity was found to be low at 4.11 UI/g Hb (normal range: 5.7–9.9 UI/g Hb). The coexistence of c.466A>G and c.1058T>C G6PDH mutations in cis confirmed the A-Betica variant associated with a moderate G6PD deficit (class III). Conclusions G6PD deficiency is variable in its clinical presentation and it should be considered in the differential diagnosis of hemolytic anemia at any age. The co-occurrence acute hemolytic anaemia and methemoglobinemia has been described in the literature infrequently.

University of Washington, USA

Iron deficiency remains the most prevalent cause of anemia worldwide but the overlap with concurrent inflammation may make its recognition challenging using traditional markers of iron status that are affected by low-grade inflammation present in common conditions, such as chronic kidney disease, cirrhosis, or heart failure. This session will present a pragmatic way of interpreting diagnostic lab tests based on available evidence of how iron deficiency manifests in conjunction with chronic inflammatory conditions, and aims to help recognize patterns associated with higher chance of response to oral or parenteral iron supplementation, and aid clinicians at making personalized decisions when patients do not fit usual guidelines.

Boston Children's Hospital

UCLA, Department of Pathology and Laboratory Medicine Los Angeles, CA, USA

Introduction: The main clinical utility of flow cytometry is immunophenotyping of hematolymphoid disorders. Here we report an unusual case in which we detected metastatic carcinoma by peripheral blood flow cytometry. A 63 year old female with history of metastatic lobular carcinoma presented with altered mental status and malignant hypercalcemia. Peripheral blood was submitted for flow cytometry due to leukocytosis (WBC 12.53 × 10³/μL) with absolute lymphocytosis (6.29 × 10³/μL). Methods: Multiparametric flow cytometry immunophenotyping is performed on the Beckman-Coulter Navios Ex Flow Cytometer with reconstituted lyophilized DuraClone antibodies against the following antigens: CD2,CD3,CD4,CD5,CD7,CD10,CD11b,CD13,CD14,CD15,CD16,CD19, CD20,CD22,CD23,CD33,CD34,CD36,CD38,CD41,CD45,CD56,CD64,CD85K,CD117,CD123, CD200,CD235a, HLA-DR,Kappa,and Lambda. Data analysis was performed in the FCS Express v7 software. A peripheral blood smear was prepared and stained by Wright's stain. A cytospin was performed on the peripheral blood and was formalin-fixed and paraffin embedded. Cytokeratin AE1/AE3 immunohistochemistry (AE1/AE3, Millipore Sigma) is performed on the cytospin specimen on Leica BOND III autostainer. Results: Flow cytometry revealed an abnormal population of CD45 negative cells, representing approximately 6% of the total cells (Fig. 1a, blue dots). They expressed CD13 (partial), CD15 (dim), and CD56, but were negative for all other markers tested. No blasts and no abnormal lymphoid population are detected. A peripheral blood smear showed leukoerythroblastosis with occasional left shifted granulocytes and nucleated erythrocytes. Scattered atypical cells are also noted with slightly irregular nuclear contour, delicate chromatin, nucleoli, and scant cytoplasm, resembling blasts (Fig. 1b). These cells stained positive for cytokeratin AE1/AE3 (Fig. 1c). Bone marrow biopsy was not performed. Conclusions Carcinocythemia, circulating carcinoma cells in the peripheral blood, is a rare phenomenon, mostly reported with metastatic breast carcinoma, and morphologically it may mimic a hematologic malignancy. CD13 (aminopeptidase N) and CD15 (Lewis x) are markers of myeloid differentiation but also expressed by epithelial cells. CD56, neural cell adhesion molecule (NCAM), is expressed in various cell types in addition to NK-cells. This case report demonstrates the utility of flow cytometry for detection of carcinocythemia, which is not previously well reported.

ALCOR Scientific, LLC Smithfield, RI, USA

Introduction: Erythrocyte sedimentation rate (ESR) is a common laboratory test used to help determine if a patient has an inflammatory condition or help monitor patients with inflammatory conditions. 48 million ESR tests are performed in the US annually. Traditional ESR testing is done using a manual or semi-automated version of the Westergren method. The iSED ELITE is the latest platform in the iSED instrument family. Results are available within 20 s with minimal hands-on time, significantly reducing time to result and improving patient care. Method: The iSED technology is based on syllectometry and utilizes photometric rheology reading to quantify the rouleaux formation, which is the earliest phase of red blood cell (RBC) sedimentation. RBCs are injected into the analyzer's flow cell, and the shear stress causes them to disaggregate. Sensors capture the difference in light transmission over time as the RBCs begin to re-aggregate. The technical innovation of the iSED analyzers consists of “directly” measuring the aggregation of the RBCs, whereas traditional methods “indirectly” measure aggregation of the RBCs by recording the length at which the cells settle in a Westergren tube. The iSED ELITE technology is highly correlated to the Westergren method, but test results are available within 20 s without the drawbacks and inconvenience of traditional ESR testing. Results: Method Correlation (iSED ELITE vs. Westergren) was performed according to CLSI Procedure H02-5. The method correlation (n = 200), analyzed by Passing Bablok regression, result in a regression equation of y = 1.0 + 1.0×, 95% CI of the slope is 0.94–1.07, 95% CI of intercept is 0.43–2.3. CuSum test is greater than 0.05, showing no deviation from linearity. Paired T-Test shows equivalence. Precision studies (40 patients X 5 replicates), across the clinical range of interest, demonstrated average precision results of 6.1% for samples less than 20 mm/hr. and 3.9%, for samples 20–60 mm/hr. Traditional ESR testing is commonly affected by environmental variables as well as variables within the sample. The iSED ELITE sample is analyzed within seconds in a closed environment, limiting potential impact of environment factors, and sampling is done by the instrument, eliminating the potential for human error and improving operator safety. Conclusion: The direct ESR measurement possible with the photometric determination of ESR based on syllectometry is comparable with the Westergren methodology, and results are available in seconds versus minutes to one hour, significantly improving patient care and laboratory efficiency.

Mayo Clinic, Phoenix, USA

In 2022, two separate classification systems for hematological malignancies were proposed: the fifth edition of the WHO Classification of Haematolymphoid Tumours (WHO-HAEM5) and the International Consensus Classification (ICC) of Mature Lymphoid Neoplasms, Myeloid Neoplasms and Acute Leukemias. Although both proposed classifications include an increased emphasis on genetic testing for the diagnosis and prognostication of lymphoid neoplasms, there are some distinctive differences. This session will review some of the key differences and similarities between the WHO-HAEM5 and ICC classifications and highlight what types of genetic testing are currently recommended for the classification of lymphoid neoplasms.

Hebei Yanda Lu Daopei Hospital Langfang, China

Background: Chimeric antigen receptor T-cell (CAR-T) therapy has resulted in remarkable efficacy for certain patients with hematological malignancies. However, it is difficult to find a high-expression biomarker with good specificity for acute myeloid leukemia (AML) due to the strong heterogeneity of AML tumor cells. Screening potential patients with AML by immunophenotyping may be an effective method for CD7-CAR-T therapy. Methods:From August 1, 2022 to November 12, 2022, of 671 AML patients tested for minimal residual disease (MRD) by flow cytometry (FCM) at the Hebei Yanda Lu Daopei Hospital, 174 were MRD positive. Among these MRD+ cases, there were 98 males and 76 females. The median age was 32 years (range: 1–76 years) and the median tumor blast proportion was 7.85% (range: 0.01%–97.83%). Biomarkers analyzed included CD34, CD117, CD45, HLA-DR, CD33, CD13, CD11b, CD56, and CD7. The CD7 expression intensity of residual normal T cells in the same sample was used as control to analyze the CD7 expression on tumor cells. Expression of tumor cells was defined as follows: Brightness as Score 5. Medium 4; Weakness 3, Partial brightness or expression 2; Partial dim 1; No expression as 0. The expression intensity and expression rate were numerically coded. Results: 1. The expression of CD7 in 174 positive AML MRD cases was as follows: 110 cases with 0 score, 5 cases with 1 score, 30 cases with 2 score, 3 cases with 3 score, 25 cases with 4 score, and 1 cases with 5 score. The overall expression rate and the rate of patients with Score ≥ 3 was 36.78% (64/174) and 16.66% (29/174), respectively. 2. CD7 expression intensity changed at different time points when follow-up. CD7 expression was gradually acquired in two patients, whose expression scores of CD7 in tumor cells was less than three previously, which was not suitable for CD7-CAR T. After treatment, the CD7 expression score reached four, could receive CAR-T therapy Fig. 2. 3. We tried CD7-CAR-T therapy in a child with AML whose tumor cells had a CD7 score of four. Complete remission was sustained for two months after the infusion of CAR-T cells, followed by bridging to allogeneic hematopoietic stem cell transplantation. Up to 3 months after transplantation, the patient had good BM regeneration with negative MRD and 100% chimeric rate. Conclusions: It is promising to select appropriate AML patients for CD7-CAR-T therapy according to analyze the CD7 expression on tumor cells.

¹Laboratory Medicine Department, Hospital Universitari Germans Trias i Pujol Badalona, Spain, ²Laboratory Medicine Department, Hospital Universitari Son Espases Palma de Mallorca, Spain

Introduction: Lymphocytosis is a frequent finding whose underlying cause may be hematologic, as in lymphoproliferative disorders (LPD), or non-hematologic, as in viral infection. These cases require smear review, in addition to frequently triggering more complex and costly tests such as flow cytometry or serology analysis. The number of publications investigating the relevance of cell population data (CPD) in differential diagnosis of lymphocytosis has increased significantly. The exploitation of these data can be a complementary tool to the manual review of the smear. Methods: 523 patients were included in the study: 316 with lymphocytosis (>5 × 10⁹/L) and/or with variant lymphocyte flag (118 mononucleosis and 198 LPD: 176 chronic lymphocytic leukemias (CLL), 12 splenic marginal zone lymphomas (SMZL) and 10 mantle cell lymphomas (MCL)) and 207 patients without alterations as controls. Leukocyte and lymphocyte count were analyzed in addition to 14 CPD parameters (mean and standard deviation of volume, conductivity and the different light scattering angles (SM, SU, SL, SA, AL2)) on the Unicel-DxH900 (Beckman Coulter). A validation cohort of 32 mononucleosis and 52 LPD was included to validate the results obtained in the initial analysis. Parameters were compared using ANOVA test and the Bonferroni-Tukey test (post-hoc analysis). ROC curves were analyzed to evaluate the diagnostic performance of the parameters and a cut-off point was selected through Youden's index. The sensitivity (S), specificity (E), positive predictive value (PPV) and negative predictive value (NPV) of the selected cut-off was calculated. Results: Most CPD parameters showed statistically significant differences between groups (table 1). Mean lymphocyte volume (MN-V-Li) was the parameter with better discriminant efficiency in LPD versus Mononucleosis (AUC = 0.882; 95% CI: 0.845–0.919; p < 0.001). A MN-V-Li ≤ 90 cut-off point was selected showing a PPV = 89.8%, NPV = 77.1%, S = 84.6% and E = 84.4% to detect LPD in the validation cohort. Conclusions: Lymphocyte volume is a candidate positional parameter to be included in the decision algorithms of lymphocytosis. The inclusion of CPD data in the decision algorithms of routine clinical practice involves the optimization of the resources available to the laboratory and may provide highly valuable objective information.

Siemens Healthineers Diagnostics Tarrytown, NY, USA

Introduction: The human body has a phenomenal ability to keep the blood in a fluid state. It perfuses all the organs of the body with needed nutrients and collects the waste products of metabolism to be taken to the kidneys for elimination. When a blood vessel is injured, blood clots plug the injury to prevent bleeding. A complex system of enzymes and proteins found in the plasma and endothelial walls of the blood vessels help to maintain this balance. Disturbances in this mechanism result in coagulation disorders. Artificial intelligence and machine learning (ML) models could identify subtle changes in routine blood markers, which could be precursors to these disturbances. Methods: We trained machine learning models with over 100 000 sets of patient data with and without coagulation disorders. The data was sourced from MIMIC-IV, a hospital-wide electronic health record (EHR) from Beth Israel Deaconess Medical Center, Boston, MA. A gradient-boosted model with 300 trees with a maximal depth of 30 layers produced optimal performance. The input parameters consisted of age, gender, and the results of routine blood markers, such as complete blood counts, differential counts, comprehensive metabolic panels, and lipid panels recorded up to 3 years before the diagnosis of coagulation disorders including von Willebrand disease, thrombophilia, antiphospholipid syndrome, activated protein C resistance, factor deficiencies, and drug-induced disorders. An evaluation of the performance was conducted using the area under the receiver operating characteristic curve (AUC). Results: We were able to show that the model can predict the risk for the listed coagulation disorders, with an average AUC of 0.89 for all the disorders combined with 0.90 sensitivity, 0.88 specificity, 0.88 positive predictive value and 0.90 negative predictive value. Neutrophils, eosinophils, monocytes, lymphocyte counts, and total bilirubin seemed predominantly to contribute to the identification of risk. Prediction accuracy was consistent up to 3 years prior to diagnosis. Conclusion: Coagulation may be enhanced by the neutrophil extracellular trap (NET)-dependent activation of the contact system. Eosinophils contribute directly by producing key coagulation factors, thrombin, and plasminogen. High bilirubin may indicate liver disease that could affect the synthesis of coagulation factors and compounds of the fibrinolytic C system in the liver, leading to bleeding diathesis. Thus AI/ML-based prediction models may be able to help identify the risk for coagulation disorders years before symptoms appear, such as bleeding or intravascular clotting.

Geisinger

Viscoelastic assays are global tests of hemostasis performed on whole blood. Although previously marketed as point of care devices, technological limitations such as manual pipetting and liquid quality control, limited their implementation. Recently cartilage-based testing emerged onto the market, making the technology more widely available. Unfortunately, most of the guidelines for viscoelastography are clinically based with an emphasis on interpretation and assessment of hemostasis in different disease states as well as the utilization of blood products based on the results. The objectives of the session will focus on providing guidance for successful implementation of viscoelastrography including selection criteria for devices, staff qualifications, training and competency, safety and waste disposal, information technology assessment and integration, and quality control/assurance requirements and framework.

University of Washington

Transgender and non-binary people have a gender identity that differs from their sex assigned at birth. Standard of care for these individuals is to administer gender affirming hormones to promote development of secondary sex characteristics that align with gender identity. A masculine-identified person will take testosterone; a feminine-identified person will take estradiol with or without anti-androgens. Testosterone stimulates erythropoiesis, leading to different reference intervals for red cell indices between adult cisgender men and women. Estrogen treatment in cisgender women is associated with procoagulant profile. This session will highlight the current knowledge related to interpreting hematology and coagulation results in gender diverse people prescribed gender affirming hormones.

Institut fur Transfusionsmedizin, UniversitÃtsmedizin Greifswald, Germany

UniversitÃtsmedizin Greifswald Greifswald, Germany

Platelet factor 4 (PF4, synonym: CXCL4) is an evolutionary old chemokine with proposed roles in hemostasis and antimicrobial defense. In addition, PF4 has attracted considerable attention as a crucial mediator of one of the most prothrombotic adverse drug effects affecting blood cells, heparin-induced thrombocytopenia (HIT). Interest in PF4 substantially increased in 2021 when it was identified as the target antigen in the life-threatening adverse effect, vaccine-induced immune thrombotic thrombocytopenia (VITT). The lecture will address the concept that a major biological function of PF4—a strongly cationic chemokine—is to bind to negatively-charged prokaryotic microorganisms. This results in structural changes in PF4 that trigger a danger signal recognized by the adaptive immune system. Application of biophysical tools has provided substantial insights into the molecular mechanisms by which PF4 becomes immunogenic, providing insights into a new mechanism of autoimmunity. According to this new mechanism, binding of autoantibodies with high affinity induces conformational change(s) in endogenous proteins. When high avidity anti-PF4 autoantibodies bind to PF4, it is recognized as foreign antigen, which leads to the prothrombotic disorders, autoimmune HIT and VITT. Both present prototype examples of antibody mediated thrombo-inflammation and immune-thrombosis. Immune complexes containing PF4 and anti-PF4 IgG cluster and signal through FcRgIIA, which generates procoagulant platelets, activates monocytes, induces platelet/neutrophil aggregates, and stimulates Netosis by neutrophils. DNA released by Netosis amplifies immune injury and activates complement, which deposits on the endothelium. Endothelial cells become activated, expressing tissue factor and releasing von Willebrand factor. Von Willebrand factor binds PF4 and subsequently anti-PF4 antibodies, which in turn further activates neutrophils and further propagates thrombin generation. These self-enhancing amplification loops can be interrupted by inhibiting thrombin (generation) by non-heparin anticoagulants and further upstream by blocking FcRgIIA by high dose IVIG. Current assays for HIT and VITT show how structural aspects of anti-PF4 pathobiology relate to assay design and performance characteristics. Currently, functional (platelet activation) assays using washed platelets detect HIT antibodies when heparin is added, and VITT antibodies when PF4 is added. Solid-phase PF4-dependent immunoassays using microtiter plates are sensitive for both HIT and VITT antibodies. However, rapid immunoassays, in which the PF4/heparin antigen is coated on beads, are sensitive and specific for HIT, but not for VITT antibodies. Reports are evolving describing an association of PF4-dependent platelet activating antibodies in patients with (recurrent) venous and arterial thromboses. These reports might indicate that such anti-PF4-antibodies could be of clinical relevance beyond HIT and VITT.

¹Laboratorio de HemocitologÃa, Hospital Fernández Buenos Aires Argentina, ²Laboratorio de HemocitologÃa, Hospital Fernández Buenos Aires Argentina, ³Laboratorio de HemocitologÃa, Hospital Fernández Buenos Aires Argentina, ⁴Laboratorio de HemocitologÃa, Hospital Fernández Buenos Aires Argentina, ⁵Laboratorio de HemocitologÃa, Hospital Fernández Buenos Aires Argentina, ⁶Laboratorio de HemocitologÃa, Hospital Fernández Buenos Aires Argentina

Introduction: Circulating erythroblasts are precursors of red blood cells confined, in adults, to the bone marrow. In fetuses and neonates, these cells can be found in low concentrations in peripheral blood. The appearance of erythroblasts in peripheral blood is a marker of abnormal erythropoiesis and/or hypoxic stress. Several studies demonstrate that the evolution of the count of these cells could serve as an early indicator of mortality. Since erythroblasts are nucleated cells, they interfere with the measurement of white blood cells, which is why having an automated hematology analyzer that has a specific chamber for its detection and counting is extremely important. The objective of this study was to compare the NRBC count obtained by the Mindray BC-6200 hematology analyzer versus that obtained by optical microscopy. Methods In order to compare the two methods and thus determine whether the automated count could replace the manual count with sufficient precision, we used Spearman's correlation coefficient and the bias obtained by the Bland-Altman method. 145 samples from patients at Hospital Fernández (Buenos Aires, Argentina) were processed, the NRBC count was performed by two methods: automated in the BC-6200 hematology counter and manual by microscopy in peripheral blood smears. To evaluate the manual NRBC counts, all samples included in this study underwent microscopic review according to the morphological criteria defined by the International Council for Standardization in Hematology (ICSH) recommendations. For the statistical treatment of the data, the Graph Pad Prism program was used. Spearman's correlation coefficient and concordance by the Bland Altman method were analyzed. Results In this way, not only the linear association between the measurements is evaluated, but also the concordance between them. 145 samples from patients at Hospital Fernandez covering a measurement range from 0 to 800 NRBC/100 WBC (linearity reported by the manufacturer: 0–600 NRBC/100 WBC) were analyzed, obtaining a Spearman correlation coefficient of 0.9682 was obtained and bias obtained by the Bland-Altman method was −0.74. Conclusions The possibility of automating the NRBC count implies a benefit, especially in those leukopenia patients where it is difficult to perform the count by optical microscopy. With this results we can say that we could use the quantification of NRBC by the Mindray BC-6200 with excellent performance.

Laboratorio de HemocitologÃa, Hospital Fernández Buenos Aires, Argentina

Introduction: More than 1% of morphologically identified schistocytes on the blood film are considered suspicious for thrombotic microangiopathy (TMA) ^[1]. Schistocytes should be identified by experienced morphologist. The fragment red cell count (FRC) generated by some automated hematological analyzers provides a valuable screening tool for the presence of schistocytes^[1]. The Mindray BC-6200 hematology analyzer implements the adaptive multidimensional morphological boundary classification technology to detect FRC. which would be an extremely useful tool in the suspicion of morphological alterations of erythrocytes. The objective of this work was to analyze the behavior of FRC% and FRC# in samples with and without morphological alterations of the erythrocytes. Methods We processed 294 samples in BC-6200 hematology analyzer from patients at Hospital Fernández, including 77 samples without morphology abnormalities. 120 with poikilocytosis but without schistocytes /spherocytes and 97 with schistocytes and spherocytes; the morphology of erythrocytes was reviewed by optical microscopy in peripheral blood smears. For the statistical treatment of the data, the Graph Pad Prism program was used. The results were expressed as median and maximum and minimum. To assess the significance of the differences between the groups, the Kruskal-Wallis test was used with Dunn's post-test of multiple comparisons when the differences were significant. Results The values obtained were: Without morphological abnormalities Poikilocytosis Schistocytes + Spherocytes FRC % 0.2 (0.10–0.82) 0.66 (0.20–2.37) 1.20 (0.24–18.20) FRC # 0.0088 (0.0045–0.0376) 0.02605 (0.0072–0.0817) 0.0387 (0.0078–0.2940) Significant differences (p < 0.05) were observed in the FRC% and FRC# parameters among the three groups analyzed. Conclusions We observed an increase in FRC% and FRC# in samples that present poikilocytosis, and even more in those with schistocytes and spherocytes, this is due to the reduction of non-specific binding of proteins on the cell membrane to fluorescent dyes, so the fluorescent signal of RBC fragments is smaller than that of intact erythrocytes (due to the size and the content of hemoglobin). It is proposed that these parameters should be taken into account when analyzing blood samples and could be used as smear review criteria.

Reference

[1] Zini G, 2021 update of the 2012 ICSH Recommendations for identification, diagnostic value, and quantitation of schistocytes: Impact and revisions. Int J Lab Hematol. 2021;43(6):1264-1271.

¹National Institute of Blood Disease Karachi, Pakistan, ²The Sahara University Narowal, Pakistan, ³NED University of Information and Technology Karachi, Pakistan

The word ‘CBC analyzer' and ‘Hematology analyzer' are alternatively practiced in diagnostics, is it just a matter of misnomer practices? Trust me! Isn't. Again, the scope of ‘CBC testing' and ‘hematology testing' can be replaceable to each other, is it just a matter of misinterpret-practices? Trust me! It is. Here, this trust of 'Isn't' and 'It's' is largely built after the introduction of extended analytical channels in modern CBC analyzers. These extended analytical channels are now generating the real-time ‘3- Dimensional (3-D)' details of normal but immature and or abnormal blood cells, too. The referred 3-D details include the counts and cellular contents (concentration) are de-facto qualitative and quantitative information, respectively. What's-more, the qualitative information is for-real the CBC-based automated morphometric parameters. Focusing more than a hundred quantitative and qualitative parameters (high-dimensionality data) generated by modern hematology analyzers like XN-1000 (Sysmex) with routine CBC analysis, is it a ‘dare' or a ‘challenge'? Here, the dare of this high-dimensionality data may be a large window of their deviational patterns (fingerprints) for any particular disorder, while a challenge will be of identification, interpretation, and visualization of the disease's fingerprints. The indicated fingerprinting and then recalling of matching patterns from all combos (library) for test samples, is nearly beyond the bounds of possibility for any human brain (intelligence) but not for the artificial brain (machine). In present study, the steps of finding patterns and their recalling were address through machine learning models through computer command language (Python). Where various models were applied by the distribution of 70:30 for training:testing on total 5989 CBC analysis belonged to 65 disease groups. In results, worth discussing points including a noticeable accuracy of 85.61% with 91.92% precision for Random Forest Classifier (RFC) followed by Decision Tree (DT), Support Vector Machine (SVM), Logistic Regression (LR), K-Nearest Neighbor (KNN), Stochastic Gradient Descent (SGD), and Gaussian Naive Bayes (GNB) were noted. The website was integrated with model file of our best predictive classifier (RFC). Local hosting with live demonstration option through premium of Heroku was set-upped for website, where option drag-and-drop option is added along cop-and-past of CBC output file. On data feeding the prediction for any test case come-up with various easy to understandable visualization and metrics. Here, Well said by Dr. Keith: "AI is not going to replace physicians, but physicians who use AI are going to replace physicians who don't, and that may be the cautionary tale".

Leiden University Medical Center, The Netherlands

Standardization of hemoglobin A2 for beta-thalassemia screening; update by the IFCC/ICSH Joined Working Group.

Cornelis L. Harteveld (on behalf of the JWG)

Standardization of laboratory tests is instrumental for the detection of beta-thalassemia carriers when using HbA2 levels as a first indicator. A variety of HPLC and CE instruments are on the marked to separate Hb fractions and quantitate the HbA2 and used in screening- and diagnostic laboratories as part of a premarital, preconception or antenatal screening program for the prevention of hemoglobinopathies identifying beta-thalassemia carriers by an elevated HbA2. The present WHO HbA2 standard is already over 30 years old and reaching its end. The IFCC standardization project started in 2004 and in 2015 the ICSH joined in collaboration in the IFCC/ICSH Joined Working Group (JWG). The project involves the developing a reference measurement procedure, defining and producing calibrators and certified reference materials. Recently a road map was published to illustrate the steps still to be taken and the role of the different stakeholders in this process.

An overview will be given of what has been done so far, and what still needs to be done to achieve the final goal, that is, a certified reference material for HbA2 measurement to facilitate standardization worldwide.

¹Division of Clinical Laboratory, Tottori University Hospital Yonago Japan, ²Department of Pathology, Tottori University Hospital Yonago Japan, ³Division of Clinical Laboratory Medicine, Department of Multidisciplinary Internal Medicine, School of Medicine, Tottori University Faculty of Medicine Yonago Japan

[Background] Although evaluation of bone marrow cellularity is largely dependent upon visual estimates, it is rapid but roughly semi-quantitative. The aim of this study was to construct an automatic quantification method for bone marrow cellularity using image analysis software. [Methods] We used HE specimens of bone marrow biopsies and clots from patients who underwent bone marrow examination at Tottori University Hospital from 2020 to 2022. Cellsens (OLYMPUS) and WinRoof2015 (Mitani Corporation) were used for image analysis. We developed a method in which the non-fat area is used as a quantification value for cellularity (method A), and a method in which the nuclear area of cells is quantified by discriminant analysis after monochromeizing the image and the cellularity is calculated (method B). The cellularity evaluation was compared with the visual estimates in pathology reports. [Results] A total of 91 HE specimens were used in 54 cases. The male to female ratio was 29:25, 38 biopsy and 53 clot specimens were included. Cellularity was visually scored as hypocellular (N = 17), normocellular (N = 44), or hypercellular (N = 30). In comparison with the visual estimates in the pathology reports, the correlation coefficients of methods A and B were 0.80 and 0.85, respectively. In ROC analysis with hypocellularity as the objective variable, the AUC for method B was significantly higher (0.95 vs. 0.88, p < 0.01). In ROC analysis with hypercellularity as the objective variable, AUC of method B showed higher than method A (0.96 vs. 0.93, p = 0.18). Akaike's information criterion (AIC) of methods A and B were 495.3 and 504.9, respectively. Next, we developed a method C by combining methods A and B. When comparing the method C with the visual estimates in the pathology reports, the correlation coefficient was 0.88, and the AUCs for hypocellularity and hyper cellularity were 0.94 and 0.97, respectively. In addition, AIC was 467.4, which was the lowest among the three methods. [Conclusions] The quantification value by method A was overestimated in hypocellular marrows and that of method B tends to overestimate in hypercellular marrows. With the method C, the most appropriate values can be obtained for the pathology reports.

McMaster University, Canada

Introduction: Coagulation factors, anticoagulants, and fibrinolytic proteins are important for hemostasis, and mutations affecting these proteins cause some rare inherited bleeding disorders that are particularly challenging to diagnose. This review provides current information on rare inherited bleeding disorders that are difficult to diagnose. Methods: A review of the literature was conducted for up-to-date information on rare and difficult to diagnose bleeding disorders. Results: Some rare bleeding disorders cause an inherited deficiency of multiple coagulation factors (F), such as combined FV and FVIII deficiency and familial vitamin K dependent clotting factor deficiency. Additionally, congenital disorders of glycosylation can affect a variety of procoagulant and anticoagulant proteins and also platelets. Some bleeding disorders reflect mutations with unique impairments in the procoagulant/anticoagulant balance, including those caused by F5 mutations that secondarily increase the plasma levels of tissue factor pathway inhibitor as well as THBD mutations that increase functional thrombomodulin in plasma or cause a consumptive coagulopathy due to thrombomodulin deficiency. Some bleeding disorders accelerate fibrinolysis due to loss-of-function mutations in SERPINE1 or SERPINF2, or in the case of Quebec platelet disorder, a duplication mutation that rewires PLAU and selectively increases expression in megakaryocytes, resulting in a unique platelet-dependent gain-of-function defect in fibrinolysis. Conclusions: Current information on rare and difficult to diagnose bleeding disorders indicates they have unique clinical and laboratory features, and pathogenic characteristics to consider for diagnostic evaluation. Laboratories and clinicians should consider these disorders in their strategy for diagnosing bleeding disorders.

¹Department of Clinical Laboratory Medicine, Juntendo University Graduate School of Medicine Tokyo Japan, ²Department of Advanced Research Institute for Health Science, Juntendo University Graduate School of Medicine Tokyo Japan, ³Sysmex Corporation Kobe Japan

Introduction: We have constructed a Deep Learning System (DLS) that uses Convolutional Neural Networks (CNN) to recognize cell characteristics by cell classifications and morphological abnormality comments in peripheral blood (PB) smears. In this study, the cell classification performance and ability to detect abnormal cells by this system was evaluated using routine PB smears. MATERIALS AND Methods: Training of the morphological DLS was performed by CNN using approximately 1 million digitalized PB cell images collected at Juntendo University Hospital (Tokyo, Japan). PB smears were stained with May Grunwald-Giemsa using a fully automated slide-maker SP-50 (Sysmex), and the digital cell images were obtained by DI-60 (Sysmex). The performance of DLS was evaluated using total 139 378 images from 589 PB smears (375 patients and 214 healthy controls). Data were analyzed using JMP15 software (SAS). Results: The accuracy of cell classification ability was 97.3% or higher for all 17 classifications. The accuracy of assigning abnormal comments was 90% or higher for 11 of 12 abnormal comments related to lymphoid cells in the lymphoma cases and 12 of 16 abnormal comments related to mature neutrophils in the MDS cases. In all 118 PB smears determined to have abnormal cells by routine microscopic examination in the hematology laboratory, the DLS was able to detect abnormal cells with the classifications to blasts (n = 115) or morphological abnormality of lymphocytes (n = 3). Abnormal comments of lymphocytes appeared more frequently in the lymphoma group than in the HC group (p > 0.05). CONCLUSIONS The constructed DLS showed good performance not only in cell identification, but also in the recognition of morphological abnormalities that contribute to the detection of abnormal cells.

Pro Dr Holly Makkah, Saudi Arabia

Background: Treatment of chronic myeloid leukemia had been revolutionized by the use of Tyrosine kinase inhibitors. Recently, the strategy of treatment free remission had been introduced to reduce toxicity and financial burden beside allowing patients who now have similar survival to normal people to achieve some goals as childbearing. Aim of this study: to check the suitability of imatinib as a first line treatment in potential patients in whom treatment free remission will be an ultimate goal. Methods: This study included patients with chronic myeloid leukemia, chronic phase in the period between 2011 and 2020. Initial diagnosis was done by bone marrow, cytogenetics, molecular testing for BCR/ABL in addition to quantitative estimation of BCR-ABL1 on peripheral blood every 3 months. Results: 140 patients were included in this study who received imatinib as first line therapy. The male to female ratio was 1.1 and the median age was 42 years. The majority of patients were considered low risk (70.2%) as by EUTOS risk score. Breakpoint region was e13/a2 in 39.3%, e14/a2 in 38.3% and unknown in 22.7% of patients. 62% of patients were compliant with treatment. At the time of analysis, 79 patients (56%) remained on imatinib. 51/140 (36%) stopped it due to failure, 11 due to intolerance (7.9%). Favorable molecular responses were achieved in 67.4%, 53.9% and 50% at 3, 6, 12 months respectively Compliance and risk scores were significantly affecting these responses. 54/70 (77%) of those who achieved MMR at 12 months achieved a deep molecular response (M4, M4.5) after a median period of 36 m. 40 patients (57%, 74%) of those who achieved MMR, DMR respectively developed DMR lasting for more than 2 years (which is the requirement for TTF). This also reflects 28.5% of the whole group of patients. Conclusions: Although Imatinib continues to be the preferred first line in treating CML, CP yet it might be more suitable to use a second generation TKI if TFR is planned ahead.

Professor Dr Nahla Ahmed Holloy Makkah, Saudi Arabia

Case presentation: A 51-year-old male patient, a heavy smoker, was diagnosed as stage IV poorly differentiated adenocarcinoma of the lung in the 15 December 2019. Anaplastic lymphoma kinase (ALK), Cros oncogene 1 (ROS-1), and Epidermal growth factor receptor (EGFR) mutations are undetectable but Programmed death ligand 1 (PDL-1) is 1% expression. The patient received the first cycle of chemotherapy plus pembrolizumab in the 6 January 2020. A total of three cycles of Carboplatin plus Pemetrexed and Pembrolizumab, then Pemetrexed and Pembrolizumab as maintenance therapy. The disease was well controlled with excellentpe rformance status since then for almost one year and half. During the treatment course, Pemetrexed was dropped due to the gradual decrease of (GFR).Before the current admission, the last maintenance dose of pembrolizumab was in the 19 December 2021. Before the emergency visit, he began to have fever, diarrhea, abdominal pain, and vomiting for 5 days. He was denying eating from outside and contacting sick individuals. Systemic review was unremarkable for neurological, cardiovascular, and respiratory symptoms. He was evaluated in the emergency department and physical examination was normal apart from fever of 38.5 degrees Celsius. Peripheral blood smear showed normochromic anemia, anisocytosis, rare nucleated red cells, and rare fragmented red cells. The red cell indices were; Hemoglobin level 5.8 g/dL, RBC 1.8 × 10¹²/L, MCV 96.8 fL, PLT141 × 10⁹/L. White blood cells showed Leukocytosis: 45.8 × 10⁹/L. The leukocyte differential count revealed sever neutropenia: neutrophils 4%, lymphocytes 3%, and blasts 93%. Coagulation studies: Prothrombin time (PT) and partial thromboplastin time (PTT) were within the normal range: PT17.1 sec. (RR 11–16), INR 1.3 (RR 0.8–1.2), PTT 31.60 sec. (RR 28–44). However, fibrinogen and D-Dimer were elevated: 6.6 g/L (RR 2.0–4.0) and 4.90 (High) μg/mL (RR < 0.55) respectively. Chemical studies: creatinine level 5 mg/dL. Patient started on broad-spectrum antibiotic IV fluid and admitted to hospital. Further tests were performed included cultures (blood, urine, stool, and stool examination). In addition, Bone marrow (BM) studies were done which revealed: Bone marrow aspirate smears were particulate with cellular particles and trails, depressed trilineage hematopoietic series. The blasts were increased; around 80%–85%, with high N/C ratio, basophilic granulated cytoplasm, round/ oval nuclei, occasional Auer rods, some blasts show prominent nucleoli. The BM biopsy revealed 90% cellularity with interstitial involvement by immature looking mononuclear cells; blasts in a background of trilineage hematopoietic series. (Fig. 1).Cytogenetic results revealed a normal chromosomal count (46, XY [20]) without apparent clonal abnormalities. In addition, Fluorescence in site hyperdization (FISH) did not reveal any clonal abnormalities. Molecular studies revealed negative BCR-ABL (p210, p190) and PML-RARA. While FLT3 (ITD) and NPM1 mutations were positive. Over the next days, during admission, the patient developed lung aspergillosis and started on antifungal treatment. Unfortunately, the patient was not fit for induction treatment of acute myeloid leukemia and referred for best supportive care.

Khoo Teck Puat Hospital Singapore, Singapore

Introduction: Thrombocytopenia is a condition characterised by an abnormally low platelet count. The causes of thrombocytopenia are broadly classified into two categories: ineffective platelet production and increased platelet destruction or consumption. In Khoo Teck Puat Hospital (KTPH), the adult reference interval for platelet count is 173 to 414 × 109 cells/L. Immature platelet fraction (IPF) measurement is available on the haematology fluorescence flow cytometry analyser Sysmex XN-9000 (Sysmex, Japan) which is used in the Department of Laboratory Medicine. IPF is suggested to be useful in the differential diagnosis for thrombocytopenia. The authors seek to perform a clinical validation on the utility of the IPF parameter in an acute care hospital in Singapore. Methods: IFP results from fifty thrombocytopenic patients were correlated against their clinical diagnosis. Fourteen patients had undergone bone marrow aspirate procedure for clinical diagnosis. Three patients also tested on A Disintegrin-like Metalloprotease domain with ThromboSpondin type 1 motifs (ADAMSTS13) assay to confirm thrombotic thrombocytopenic purpura (TTP). Dengue serology tests (dengue NS1 antigen, Dengue IgM and Dengue IgG) were performed on thirty-three patients to confirm dengue virus infection. The patients are grouped according to the two manufacturer recommended clinical ranges: 1.1%–6.1% which is indicative of ineffective platelet production and above 6.2% which is indicative of increased platelet destruction or consumption. Sensitivity, specificity and predictive values of IPF was calculated with data collected. Results: In patients with conditions resulting in ineffective platelet production, the IPF parameter showed a sensitivity and specificity of 100% and 76% respectively. The positive and negative predictive values are 27% and 100%. The overall efficiency is 78%. In patients with conditions resulting in increased platelet destruction or consumption, the IPF parameter showed a sensitivity and specificity of 76% and 100% respectively. The positive and negative predictive values are 100% and 27%. The overall efficiency is 78%. Conclusions: The study shows that the IPF parameter is useful in the differential diagnosis of thrombocytopenic patients. The authors suggest that the inclusion of IPF parameter as a reflex test for patients with unclear or unknown cause of thrombocytopenia may provide insight on additional diagnostic tests that can be requested for confirmation of diagnosis in patients with thrombocytopenia.

MD Anderson

McMaster University, Canada

An increasing number of machine learning applications are being developed and applied to digital pathology, including hematopathology. The goal of these modern computerized tools is to support diagnostic workflows by extracting and summarizing information from multiple data sources, including whole-slide images (WSI) of human tissue. Hematopathology is inherently multimodal and can serve as an ideal case study for machine learning applications in pathology. However, hematopathology also poses unique challenges when applying machine learning approaches. By modeling the pathologist workflow and thinking process, machine learning algorithms can be designed to address practical and tangible problems in hematopathology. In this presentation, the role of machine learning in hematopathology workflows will be discussed. Currently machine learning research trends in the field will be reviewed, with a focus on bone marrow cytology and histopathology, and how adoption of new machine learning tools may be enabled through the transition to digital pathology. Finally, potential future approaches such as multimodal representations will be reviewed, including how such approaches may one day serve as clinical decision support tools in hematopathology.

¹Diagnostic Laboratory Sciences and Technology Research Center, School of Paramedical Sciences, Shiraz University of Medical Sciences, Shiraz, Iran, ²Peyvand Pathobiology and Genetic Lab Complex, Shiraz, Iran

Introduction: Lysosomal storage disorders (LSDs) are a group of rare inherited diseases, which can go undiagnosed for years due to their heterogeneity. Galactosialidosis is a rare form of LSDs caused by mutations affecting CTSA gene, that encodes the protective protein/cathepsin A. It results in deficiencies of β-galactosidase and neuraminidase 1. Clinical manifestations of galactosialidosis is similar to other lysosomal disorders and its definitive diagnosis is challenging due to lack of available diagnostic test and its infrequency. Methods and results: A 7-year-old female presented with mucopolysaccharidosis-like symptoms including rough facial features, cardiomegaly, and intellectual disability. The size of liver and spleen were normal with no mass lesion. Except for mild neutropenia (ANC: 1.75 × 10⁹/L), no other abnormal hematologic parameters were found in the initial investigation. Comprehensive metabolic and thyroid panels also revealed no noticeable abnormalities in kidney, liver, or thyroid function (TSH: 1.75 IU/mL; T4: 9.5 μg/dL; BUN: 17 mg/dL; Cr: 0.38 mg/dL). No atypical Alder granulation was found in granulocytes and the level of glycosaminoglycansin urine was normal (urine-MPS; 4.2 mg/mmol creatinine), which almost exclude the diagnosis of mucopolysaccharidosis. However, the prominent cytoplasmic vacuoles in peripheral blood smear lymphocytes were still suggestive of a storage disorder. Accordingly, whole exome sequencing was proposed as the next lab investigation and diagnosis of galactosialidosis was confirmed by homozygous mutations in the CTSA gene (c.556C>A: p.P186T). Conclusion: Unlike the cytoplasmic vacuoles in neutrophils and monocytes, which could form in different conditions due to their phagocytic characterization, the presence of these vacuoles in lymphocytes should not be simply ignored as they could be leading evidence for diagnosis of an LSD.

¹West China Hospital of Sichuan University Chengdu China, ²The First Bethune Hospital of Jilin University Jilin China, ³The First Affiliated Hospital of Xi'an Jiaotong University Xian China, ⁴The First Affiliated Hospital of Soochow University Suzhou China, ⁵Huashan Hospital Fudan University Shanghai China, ⁶The First Affiliated Hospital, Sun Yat-sen University Guangzhou China, ⁷Chinese PLA Ceneral Hospital Beijing China, ⁸Peking University First Hospital Beijing China, ⁹Tongji Medical College of Hust Wuhan China, ¹⁰Nanfang Hospital Guangzhou China, ¹¹Ruijin Hospital, Shanghai Jiaotong University School of Medicine Shanghai China

Background: Over the past decades, the demand for digital morphology analyzers has increased because conventional microscopic examination is labor-intensive, time-consuming, and requires professionally trained personnel. ICSH recommended the hematology laboratories perform comparison studies before utilizing the digital imaging. The Mindray MC series is a new artificial intelligent-based digital morphology analyzer, and a comprehensive evaluation is essential to use it confidently in laboratories. This study investigated the performance of the instrument in a multicentric setting involving eleven tertiary hospitals in China. Methods: One hundred normal and five hundred abnormal blood samples were analyzed by the MC series and then verified by the morphologists for WBC differential. The pre-classification performance was evaluated by comparing the results before and after verification. The same blood samples were also reviewed by manual microscopies according to the CLSI H20-A2 guidelines, and the post-classification results were correlated with the manual differential to evaluate the post-classification performance. RBC classification function was also studied with a similar procedure. Blood samples with different PLT levels were also analyzed by the MC series, and the PLT counts were compared with that measured by the PLT-O mode of BC-6800plus. We also collected eight EDTA-PTCP blood samples in both EDTA-K2 and sodium citrate-containing vacutainers and then compared their PLT clump numbers with PLT-pro mode. The within-run imprecision and blood smear reading efficiency were also investigated. Results: The agreements between MC series pre-classification and verification on five types of normal WBCs were all very high (κ > 0.96), and they were lower but still acceptable for abnormal WBCs classification (κ = 0.90). The WBC post-classification results correlated very well with manual microscopy review for normal WBCs (r > 0.94) except for basophils (r = 0.848). The regression coefficients for immature granulocytes, blast cells, and nucleated RBCs were above 0.94. For RBC classification, the clinical sensitivity and specificity of each RBC abnormality were all above 90% compared with manual microscopy reviews after verification. The PLT counts estimated by the digital morphology analyzers before and after verification both correlated very well with those measured by the PLT-O approach (r = 0.98). The PLT clumps were classified and counted in the EDTA-PTCP blood samples, and significantly more PLT clumps were detected in the EDTA-K2 group than in the sodium citrate group (p < 0.05). Conclusion: The MC series is an accurate and reliable digital cell morphology analyzer and offers another intelligent option for hematology laboratories.

¹Special coagulation, Stanford Health Care Stanford, CA, USA, ²Department of Pathology, School of Medicine, Stanford University Stanford, CA, USA

Introduction: PF4-dependent enzyme-linked immunoassay (PF4/IgG) is widely used to screen for Heparin-induced thrombocytopenia (HIT) due to its high sensitivity and optical density (OD) predicting the likelihood of HIT but with disadvantages: high false-positive rate and time-consuming that requires batching samples and not realistic for urgent investigation. In our study comprising majority of positive PF4/IgG patients, we compared the performance of PF4/IgG with a rapid, on-demand, fully automated, chemiluminescence-based immunoassay by HemosIL® HIT-Ab_(PF4/H) and defined HIT-Ab_(PF4/H) titer-specific HIT probability. Methods: 46 consecutive patients from Jan. to Dec. 2022 with HIT antibodies tested by both PF4/IgG and HIT-Ab_(PF4/H), HIT function tested by whole-blood impedance aggregometry (WBIA) and 4Ts score (4Ts) provided by hematologists were included. Patients with a positive PF4/IgG (OD > 0.4) and WBIA were considered HIT-positive; others were designated HIT-negative. Results: Of total 46 patients, 40 (86.9%) were PF4/IgG positive and 19 (41.3%, 11 positive and 8 negative) had additional serotonin release assay (SRA) performed showing 100% agreement with WBIA. PF4/IgG and HIT-Ab_(PF4/H) had an 84.8% agreement, both with 100% sensitivity. HIT-Ab_(PF4/H) showed a higher specificity (54.2% vs. 25.0%, p < 0.05) with fewer false-positive cases (11 vs. 18 of 46). Majority of the fewer false-positive cases (5/7, 71.4%) occurred after cardiovascular surgery. Furthermore, these 46 patients were divided by 4Ts "0–3", "4–5" and "6–8" with 15 (32.6%), 17 (37.0%) and 14 (30.4%) patients in each corresponding group. Under “4Ts (0–3)” group, all 15 patients were WBIA/HIT negative with low titer of HIT-Ab_(PF4/H) (mean: 0.6, median: 0.3, range: 0.0–2.5 U/mL) except for one (6.6 U/mL; negative by both WBIA and SRA). Significantly increased HIT-Ab_(PF4/H) (mean: 8.8, median: 4.2, range: 0.1–66.6 U/mL) was observed in “4Ts (4–5)” group with 47.1% (8/17) HIT-positive. All 14 patients in “4Ts score (6–8)” were true-positive by both PF4/IgG and HIT-Ab_(PF4/H), associated with the highest HIT-Ab_(PF4/H) (mean: 41.5, median: 21.4, range: 8.0 ≥128.0 U/mL). Overall, HIT probability was 100.0% (10/10) with HIT-Ab_(PF4/H) result ≥20.00 U/mL (strong-positive), 76.9% (10/13) with result of 5.00–19.99 U/mL (positive) and down to 20% (2/10) for a weak-positive result (1.00–4.99 U/mL). Conclusions: In this cohort, HIT-Ab_(PF4/H) (cut-off: 1.00 U/mL) showed 100% sensitivity and a better specificity than PF4/IgG with less false-positive observed among patients after cardiovascular surgery, potentially benefiting this population. Higher HIT-Ab_(PF4/H) result showed greater likelihood of HIT, consistent with published literature on larger-scale studies. Given its rapid, automated and on-demand test capabilities, HIT-Ab_(PF4/H) facilitates a timely and reliable screen assay for HIT.

¹Special Coagulation, Stanford Health Care Stanford, CA, USA, ²Department of Pathology, School of Medicine, Stanford University Stanford, CA, USA

Introduction: Although Factor XIII (FXIII) deficiency (F13D) is a rare bleeding disorder, the actual incidence of both congenital and acquired F13D may be underreported. This may be due to diagnostic test variability and the widespread clinical laboratory reliance on urea clot solubility testing, which cannot detect mild-to-moderate F13D. HemosIL FXIII-A2 latex immunoassay provides a solution as an automated, fast, and quantitative measurement of FXIII active A-subunit for the diagnosis of F13D. Objective: We evaluated the clinical relevance of F13D tested by HemosIL FXIII-A2 and set a cut-off point indicative of consequential clinical bleeding. Methods: In this retrospective single-center cohort study, we enrolled 144 consecutive patients between August 2020 and October 2022 tested by FXIII-A2. Age, ordering reason, laboratory parameters and clinical presentation were analyzed together with a decision made by experienced hematologists to determine if there was a F13D-related bleeding diathesis. A level of FXIII <70% indicates deficiency. The etiologies of acquired F13D were categorized as “Autoimmune conditions”, “Hypo-synthesis”, “Hyper-consumption”, and “unknown reasons”. Results: 51 of 144 patients (35.4%) had F13D comprising 6 (11.8%) F13D-related bleeding, 36 (70.6%) F13D-nonrelated bleeding and 9 (17.6%) no-underlying bleeding diathesis. Only one newborn (FXIII level: 37.9%) without bleeding was suspected of congenital deficiency (mother is a known F13D carrier). The etiologies of acquired F13D revealed: autoimmune conditions (3/50, 6%) including paraproteinemia, rheumatoid arthritisand solid malignancy; hypo-synthesis (14/50, 28%) arising from leukemia, infanthood, liver disease and Noonan syndrome; hyper-consumption (13/50, 26%) occurring in the context of major bleeding, surgery, or extracorporeal membrane oxygenation (ECMO) use; and the rest (20/50, 40%) having “unknown reasons”. “Unknown reasons” and Autoimmune conditions” had similar borderline abnormal FXIII levels (mean: 64.1 vs. 62.1%, median: 64.8 vs. 62.7%, p = 0.53) which were significantly higher (p < 0.001) than “hypo-synthesis” and “Hyper-consumption” groups (FXIII mean: 45.5%, 45.0%; median: 48.9%, 48.0% respectively). The six F13D related-bleeding cases (FXIII range: 21.5%–39.2%) were only seen in “hypo-synthesis” [3 infant Intracerebral hemorrhage (ICH) and 1 liver disease] and “Hyper-consumption” (2 on ECMO). 15 infants in total were enrolled in this study and all had F13D with 12 (of 15) having ICH: 5 “spontaneous” ICH associated with significantly lower FXIII compared to 7 “non-spontaneous” ICH” due to traumatic/thrombocytopenia/congenital abnormality (FXIII mean: 38.0% vs. 58.2%, median: 35.1% vs. 59.4%, p < 0.001). 3 (of 5) infants with spontaneous ICH had FXIII rechecked months later after ICH resolved and FXIII all returned to borderline abnormal (63.9%, 66.3% and 68.9%). Conclusions: Acquired F13D due to “hypo-synthesis” and “Hyper-consumption” such as in infanthood, liver disease and on ECMO could cause clinical bleeding when FXIII level <40%.

University of Washington, USA

Hemophilia A (HA) and hemophilia B (HB) are rare inherited bleeding disorders. My Life, Our Future (MLOF) was a multisector cross-sectional U.S. initiative to improve our understanding of hemophilia through widespread genotyping. Individuals with hemophilia and potential female genetic carriers were eligible to enroll at U.S. Hemophilia Treatment Centers (HTCs). Clinical genotyping was performed centrally using a next generation sequencing screen followed by validation by an orthogonal method. Clinical genotyping reports were returned to providers to return to participants. Sites provided sex and factor levels at the time of submission. Additional clinical data were abstracted from the American Thrombosis and Hemostasis Network dataset. From 2013 to 2017, 107 HTCs enrolled 11 341 subjects (68.8% male, 31.2% female) for testing for HA (n = 8976), HB (n = 2358), HA/HB (n = 3), and hemophilia not otherwise specified (n = 4). Genotyping was high yield, with reportable variants detected in most males (98.2% HA, 98.1% HB). Throughout the program, 1914 unique variants were found (1482 F8, 431 F9), of which 744 variants were novel (610 F8, 134 F9). Inhibitor data were available for 6986 subjects (5583 HA; 1403 HB). In severe HA, genotypes with the highest inhibitor rates were large deletions (77/80), complex intron 22 inversions (9/17), and no variant found (7/14). In severe HB, the highest rates were large deletions (24/42). Inhibitor risk was associated with both non-White race and Hispanic/Latino ethnicity, including a higher rate of inhibitors reported in Black (27.3%) versus White (16.2%) subjects. In summary, MLOF is the largest hemophilia genotyping project performed to date. Our results support the need for comprehensive genetic approaches in hemophilia. This effort has contributed significantly towards better understanding variation in the F8 and F9 genes in hemophilia and risks of inhibitor formation.

Central Diagnostic Laboratory, University Medical Center Utrecht, Utrecht University, Utrecht, Netherlands

Introduction: Red blood cell distribution width (RDW) is a biomarker associated with a variety of clinical outcomes. While anemia and subclinical inflammation have been posed as underlying pathophysiology, it is unclear what mechanisms underlie these associations. Hence, we aimed to unravel the mechanisms in silico by using a large clinical dataset and subsequently validate our findings in vitro. Methods: We retrieved complete blood counts (CBC) from 1 403 663 blood samples from the Utrecht Patient Oriented Database as measured on the Abbott CellDyn Sapphire, in order to model RDW by using gradient boosting regression (XGBoost) with different sets of predictors: one set with all variables (set 1), one set without percentage microcytic (pMIC) and macrocytic erythrocytes (pMAC) and mean corpuscular volume (MCV) (set 2), and one set without these variables and reticulocyte-related variables (set 3). We performed analyses in patients with and without anemia, in men and women, and in patients younger and older than 50. Additionally, we validated our model using data from the Abbott Allinity (n = 70) and data from the Abbott CellDyn Sapphire as measured in primary and secondary care (n = 513 564). We then validated our hypothesis regarding oxidative stress using an in vitro approach, where we induced blood samples with oxidative stress. Additionally, we measured erythrocyte deformability, by using Laser Optical Rotational Red Cell Analyzer techniques, and vesiculation. Results: Only pMIC and macrocytic pMAC erythrocytes and MCV were most important in modelling RDW (RMSE = 0.40, R² = 0.96), considering removing the variables in sets 2 (RMSE = 1.00, R² = 0.76) and 3 (RMSE = 1.34, R² = 0.57) resulted in significantly worse performance. Subgroup analyses showed similar performance pattern, and validation in both Allinity data (RMSE = 0.75, R² = 0.87) and primary and secondary care data (RMSE = 0.45, R² = 0.94) confirmed the generalizability of the model as built on set 1. Our in vitro result after induction of oxidative stress underscored our results, as we observed an increased RDW and decreased erythrocyte volume, as well as increased pMIC, yet no vesiculation was observed. Furthermore, we found that deformability decreased as a result of oxidative stress possibly affecting sphericity of erythrocytes. Conclusion: We found that erythrocyte size, especially pMIC, is most informative in modelling RDW, but no role for anemia or chronic subclinical inflammation. Furthermore, oxidative stress may affect deformability of erythrocytes, and may therefore play a role in the association between RDW and clinical outcomes.

¹Department of Immunology, GNA G. Gennimatas, Athens, Greece, ²Department of Laboratory Hematology, GNA G. Gennimatas, Athens, Greece, ³Â«Flowdiagnosis» Diagnostic Centre Athens, Greece, ⁴First University Department of Pediatrics, Hippokration General Hospital Thessaloniki, Greece, ⁵Hematology Clinic, GNA G. Gennimatas, Athens, Greece

Introduction: The expression of CD64 in blasts of an acute myeloid leukemia (AML) case has a critical immunophenotypic weight in favor of monocytic commitment, as described in the official guidelines of European LeukemiaNet (ELN). The same is true for CD36, which is also suggested for the investigation of erythroid commitment. The question arises about the ontogenetic origin of CD36positive (pos)/CD64negative (neg) AML. Thus, the aim of this study was to define if there are specific immunophenotypic, morphological, molecular or cytogenetic features in this type of acute leukemias (AL). Methods: We retrospectively studied the immunophenotypic, morphological, cytochemical, molecular and cytogenetic features of all CD36pos/CD64neg AL cases diagnosed during the last 12-years period in our Hospital. Results: During the above period, 16/384 (4.2%) CD36pos/CD64neg AL cases were recorded, with male patients predominating (10/16, 62.5%). Only 5/16 (31.2%) were CD34neg, of which 4/5 were CD117 strongly positive. Also, 5/16 (31.2%) were also CD117neg, but all (5/5) showed at least partial CD34 positivity, suggesting immaturity of the blast cell. Only one case (1/16, 6.2%) was HLA-DRneg, while only 2/16 (12.5%) were CD33neg. All but one cases (which had CD71strong positivity, fulfilling criteria for pure erythroid leukemia) were CD71weak, while 12/16 (75%) were MPOneg. Moreover, there were three CD38neg cases (18.8%), all of which were CD34pos (leukemic stem cell phenotype), and only 3/16 (18.8%) were CD123neg. Interestingly, a subgroup of cases (9/14, 64.3%) was distinguished with the common characteristic of a very pronounced CD25 positivity, while among six of them with an available molecular profile, 5/6 (83.3%) were FLT3-ITD mutated. Also, all nine cases appeared with type I blasts, while none of them had erythroid blast morphological features. Additionally, all nine cases were MPOneg, although 4/6 cases investigated, showed weak alpha-naphthyl-acetate esterase (ANAE) positivity. Only (2/9) cases were classified as AML-MRC (secondary to MDS and MPN, respectively), while 7/9 were classified as M0/M1. Overall, apart from the erythroleukemia case, no other case showed CD41, CD42a and CD61 positivity, while 5/7 cases investigated (71.4%) were Glycophorin-A negative. Finally, there were no significant cytogenetic findings, except from three cases with isolated findings [−7, del(5q31), 3 × 11q23 & +13, 3 × 3q14 & −7, del(7q31)]. Conclusions: Most of CD36pos/CD64neg AL cases do not match immunophenotypically with monocytic leukemias. They exhibit morphological, cytochemical and molecular heterogeneity. However, a subgroup was distinguished with abnormally increased CD25 expression in a hierarchically immature blast cell, indicating the need for prospective investigation of its ontogenetic origin.

Sysmex Corporation 4-4-4 Takatsukadai, Nishi-ku, Kobe, Japan

Introduction: The morphology of blood cells is largely influenced by the methods of smear slide preparation, especially drying conditions. Traditionally, natural drying method has been used in many western countries, while forced-air drying has been applied in some Asian countries. The former creates differential shrinking patterns depending on cell subtypes, leading to the clear discrimination between lymphocytes and monocytes, as well as the identification of hairy cells and others. In contrast, the latter creates well-elongated shapes of cells, enabling the identification of intra-cellular organelles and the evaluation of their color tones. In this study, we analyzed the cell images, using our digital morphology algorithm, and identified the differences in specific features under the two conditions. Methods: Smear slides of six normal adult peripheral blood samples were prepared under (A) forced air drying and (B) naturally drying conditions (room temperature, moisture 50%), and stained with May-Giemsa stain. The cell images were taken with DI-60 (Syodssmex Corporation), and analyzed for ninety-four parameters critical for cellular discrimination, including (1) areas of cellular and nuclear regions, N/C ratio, imaginary diameters of cells and nucleus, (2) cellular roundness (circularity), (3) color tones of nuclear and cytoplasmic areas, chromatin aggregation pattern, quantities of granules and vacuoles, and so forth. Results: The ratios of mean parameters under (A) forced-air drying to (B) natural drying conditions, were as follows (neutrophils, lymphocytes, and monocytes, respectively); (1) cellular diameter (110.7%, 115.0%, 124.0%), N/C ratio (95.2%, 91.3%, 80.6%), area of cytoplasmic region (126.5%, 193.3%, 199.1%), and (2) cellular roundness (87.6%, 93.0%, 84.3%). These indicate that forced-air drying may lead to increased cellular diameter and cytoplasmic area, as well as decreased N/C ratio and circularity, with marked increase in cytoplasmic areas of lymphocytes and monocytes as compared with that of neutrophils. Moreover, natural drying may lead to increased cellular roundness, especially in lymphocytes and monocytes. The color tones of both nucleus and cytoplasm (including chromatin and granules) were denser in (B) than in (A). Conclusions: This study realized the quantification of various blood cell image features, by using our digital morphology algorithm, and revealed the clear differences in cell image parameters between the two drying conditions. These results may contribute to the standardization of smear preparation and cell image analyses.

¹Department of Biology and the Ottawa Institute of Systems Biology, Faculty of Science, Carleton University Ottawa, ON, Canada, ²Department of Biomedical and Molecular Sciences, Faculty of Medicine, Queen's University Kingston, ON, Canada, ³School of Baccalaureate Nursing, St. Lawrence College Kingston, ON, Canada, ⁴Clinical Pathology Department, Faculty of Medicine, Mansoura University Mansoura, Egypt

Introduction: Platelet glycoprotein-1b alpha (GP1bα) is a surface glycoprotein that is functionally defective in platelet-type von Willebrand disease (PT-VWD), which is characterized by gain-of-function mutations within the β-sheet thereby converting GP1bα to an open conformation. This results in the excessive association of GP1bα with its receptor von Willebrand clotting factor (VWF), counterintuitively leading to a bleeding phenotype in PT-VWD patients. The β-sheet mutations (clustered within the Trp230–Met239 window) therefore represent an attractive site to develop peptide therapeutics. Recently, we developed the In-Silico Protein Synthesizer (InSiPS), which is a computational tool that intelligently designs small peptides to interfere with important protein-protein interactions (PPIs). Thus, we used InSiPS to generate small peptides against the GP1bα-VWF interaction. Methods: We first used InSiPS to generate a peptide candidate list specific to GP1bα. InSiPS is a genetic algorithm developed around co-occurring protein motifs that mediate PPIs. It is a closed system that generates small peptides with a given interaction profile. The resulting peptides can bind to specific regions of a target protein while also avoiding interactions with a set of predefined, non-target proteins. These peptides are constantly improved through a cycle of evaluating and mutating an initial pool of sequences. After generating a ranked-list of peptide candidates that specifically target GP1bα, we took the top five-scoring peptides and performed a combination of in-silico docking experiments and in-vitro binding assays. Results: We have generated five peptides that are specific to GP1bα. The interaction between some of these peptides and GP1bα were further confirmed in-silico using AlphaFold2. Our preliminary in-vitro data confirms the interaction between some of these peptides and GP1bα. We will further evaluate the ability of these peptides to interfere with the interaction between GP1bα and VWF. Conclusions: We conclude that InSiPS is a suitable tool to develop small peptides against pharmaceutically important targets, such as GP1bα. Future work will focus on testing the efficacy of these peptides against GP1bα in a mammalian system.

National University of Medical Sciences, Islamabad

Anemia is a condition in which the number of red blood cells or their oxygen carrying capacity is insufficient to meet the physiological needs; which may vary for the age, sex, altitude, smoking and pregnancy. Anemia is a major public health problem worldwide, especially in developing countries in Western and Central Africa and South Asia. Beside ill health, anemia is associated with poor cognitive and motor development in children and work capacity in adults, influencing economic development of country. Physical and cognitive losses due to anemia costs the developing countries up to 4.5% loss in GDP per annum.

According to WHO, about 1.8 billion people of the world are suffering from anemia and its prevalence is five times higher in low and middle income countries. Asia and Africa account for 85% of the absolute burden of anemia in high risk groups. The most vulnerable group is children below the age of 5 years and females in their reproductive years. In 2019, global anemia prevalence in women of reproductive age group (15–49 years) was 29.6% in non-pregnant and 36.5% in pregnant women. Similarly, in 2019, global anemia prevalence was 39.8% in children aged 6–59 months: the highest being in the African region (60.2%). The developing countries having the highest burden of anemia in Africa are Zambia, Mali, Congo, Nigeria, Senegal, Sierra Leone, Chad, Tonga and Burkina Faso. Similarly, among the Asian countries, the higher prevalence of anemia is seen in Yemen, Afghanistan, India, Bangladesh, Pakistan, Nepal, Burma, Kazakhstan, and Cambodia.

Anemia can be caused by nutrition-specific factors (e.g., due to insufficient intake or poor absorption of micronutrients), non-nutritional factors, or a combination of both. In developing countries, each of these factors has social determinants as well. Non-nutritional causes of anemia include malaria, worm infestation, chronic renal failure, endocrine disorders, genetic hemoglobin disorders, vitamin A deficiency and chronic inflammatory disorders. Among nutritional factors, iron deficiency is the most common cause of anemia. Iron deficiency is present in approximately a quarter to half of children aged 6–59 months and women aged 15–49 years with anemia. The lower proportion of anemia cases associated with iron deficiency is likely in populations with a high burden of anemia and infections.

Anemia prevention and control strategies include improved dietary intake and increased food diversity with increased iron bioavailability. The targeted fortification of foods for high-risk groups and iron (and folic acid) supplementation (tablets, powders, spreads) for children and adolescents are quite beneficial. However, barriers exist regarding the dissemination and uptake of these programs, including insufficient priority, lack of knowledge and education around anemia prevention, and the difficulty in meeting the needs of high-risk groups at specific times in their lives.

Although the age standardized point prevalence and Years Lived with Disability (YLD) rates due to anemia decreased from 1990 to 2019, its burden remains high, particularly in less developed countries. Despite the observed improvement in anemia, prevalence remains high in the poorest regions in the world, presenting an obstacle to reducing maternal and neonatal mortality, as well as promoting healthy early childhood development and improving school performance and work productivity. Further reductions in anemia prevalence are likely to require a combination of programs that address the infectious and nutritional determinants.

In 2012, the 65th World Health Assembly (WHA) approved commitment to halve anemia prevalence in women of reproductive age by 2025. WHO and UNICEF extended this target to 2030 to align with UN sustainable development goals (SGDs). In 2020, the prevalence of anemia in women aged 15–49 years was added as indicator 2.2.3 of SGDs. This is a global commitment but to be effective at country level a targeted campaign with regular evaluation and monitoring is required. Overcoming bureaucratic, operational, and technical barriers to anemia control will foster more effective implementation of policies sand strategies. Moreover, effectively addressing anemia in all its forms requires a firm understanding of the unique determinants of anemia in a particular setting in relation to its cultural, religious and social norms. The prevention of anemia should continue to be supported, encouraged and adopted to achieve the targets.

Department of Haematology, Singapore General Hospital Singapore, Singapore

Introduction: Myelodysplastic syndrome (MDS) is a spectrum of clonal stem cell disorders characterized by haematopoietic cell dysplasia with predisposition for leukaemic transformation. The WHO 2022 classification reorganizes MDS categories by emphasizing cytogenetics and mutations. Next generation sequencing (NGS) is a powerful tool to consolidate genetic profile of multiple biomarkers into a single test across multiple samples. With NGS incorporated into the diagnosis of MDS, we can comprehensively identify MDS-associated genetic variants involving tumor suppressor, transcription factors, signaling factors and epigenetic pathways. Methods: Diagnostic bone marrow/blood samples received between November 2021 and December 2022 from 95 patients with suspected MDS were included. The extracted DNA and RNA were amplified using the Oncomine^TM Myeloid Research assay to interrogate 40 DNA genes and 29 fusion driver genes associated with myeloid malignancies. NGS libraries were sequenced using the Ion PGM and S5 system, and the data was analyzed using in-house bioinformatics pipelines and ion reporter software. Somatic genetic alterations including point mutations, indels and fusions were consolidated, and their association with patients' demographics, bone marrow morphology, karyotyping and fluorescence in situ hybridization (FISH) findings was determined using Fisher's exact test and Wilcoxon rank-sum test. Results: The median age was 70 years (range, 24–90). A total of 143 point mutations and indels in 28 different genes, and 1 fusion transcript were identified. Mutations in TET2, TP53, U2AF1, ASXL1, DNMT3A, SF3B1, SRSF2 and RUNX1 were frequently reported. 72 out of 95 patients (75.8%) showed one or more somatic mutations and/or cytogenetic abnormalities. At least one driver mutation was identified in 69.5% of the patients. Of the 55 patients with normal cytogenetics or unavailable karyotype/FISH results, 32 had somatic variants identified by NGS. The presence of somatic mutations and the number of mutations were significantly associated with the presence of MDS-defining cytogenetic alterations (p = 0.003 and p = 0.001 respectively). No significant associations were found between the number of somatic variants and the age or gender of the patients. Conclusions MDS is genetically complex and highly heterogeneous in clinical presentation and molecular/cytogenetics profile. Integration of NGS panels in the diagnosis of MDS, in addition to morphology, aids in accurate classification, prognostication and determination of targeted therapy.

¹Department of Laboratory medicine, College of Medicine, The Catholic University of Korea Seoul, Korea, ²Internal Medicine, College of Medicine, The Catholic University of Korea Seoul, Korea

[Background] One of the tumor cell's mechanism of immune system evasion is a regulatory T cells (CD4+CD25+FoxP3), which stimulates immune tolerance and facilitate tumor progression. Increased level of CD4+CD25+FoxP3 cells in the peripheral blood is observed in various tumors such as breast, hepatocellular, ovarian and lung cancers, and is associated with poor prognosis. It was also suggested that CD4+CD25+FoxP3 cells accumulating in tumors and in the peripheral blood of cancer patients suppress immune antitumor responses and promote tumor growth. We analyzed the CD4+CD25+FoxP3 cells frequency and lymphocyte subpopulation in relation to tumor grade and prognosis of patients with gastric cancer. [Methods] The samples for the flow cytometric analysis of CD4+CD25+ populations in the whole blood of patients were prepared using the FoxP3 staining kit (Becton Dickinson, NY, USA). The CD4 T-cell, CD-8 T-cell fractions were also calculated from flow cytometric analysis using CD45, CD19, CD20, CD4 and CD8 antibodies. [Results] Total 74 patients were enrolled, and the numbers of cancer stage 2, 3 and 4 were 5, 26 and 43, respectively. Treatment response was as follows; complete response (CR) 3, partial response (PR) 36, stable disease (SD) 22 and persistent disease (PD) 13, and statistical analysis was based on two groups (CR/PR group and SD/PD group). The CD4+/CD25+FoxP3, CD8 cell, and B-cell counts showed significant differences between the treatment response group and the non-response group. CD4 T-cell, WBC count, hemoglobin, platelet count and total lymphocyte counts did not show significant differences between the two groups. In survival analysis, the number of death was 34. B-cell count (area under the curve, AUC 0.693, P-value 0.021), hemoglobin (AUC 0.700, P-value 0.018) were significantly associated with 1-year survival. [Conclusion] In this study, the level of CD4+/CD25+FoxP3 cells was significantly higher in treatment non-response group than those of response group in gastric cancer. However, it was not significantly associated with survival. Regarding survival-related factors, B cell count and hemoglobin levels were significantly associated with 1-year survival.

Department of Laboratory Medicine, College of Medicine, The Catholic University of Korea, Seoul, Korea

[Background] It is a long-standing goal to obtain accurate diagnostic clues of myelodysplastic syndrome (MDS) by detecting PB abnormalities using an automated analyzer. This study evaluated the diagnostic utility of complete blood cell (CBC) and automated cell population data (CPD) parameters of Sysmex XN analyzer (Sysmex Corporation, Kobe, Japan), especially information about morphology of neutrophils and monocytes, to screen MDS using PB [Methods] The hematology automated analyzer Sysmex XN-1500 were used and neutrophil-FSC (NEFSC), neutrophil-SSC (NESSC), neutrophil-SFL (NESFL), monocyte-FSC (MOFSC) and monocyte-SSC (MOSSC) and other CPD were analyzed. Dysplastic neutrophil was defined by microscopic findings including nuclear hypolobulation, hypersegmentation, bizarre lobulation, clumping, cytoplasmic hypogranularity and pseudo-Chediak-Higashi granules and small size. [Results] A total of 390 participants were enrolled (63 with MDS and 327 as controls; mean age, 51.6 years) during May and July 2018. The number of the sub-classification of MDS in this study were as follows: MDS with single lineage dysplasia (n = 1), MDS with multi-lineage dysplasia (n = 14), MDS with excess blasts (n = 47), and unclassifiable MDS (n = 1). In multiple regression analysis, RBC count (p = 0.0001), platelet count (p = 0.0001), NEFSC (p = 0.0002), NESSC (p = 0.0138), QFlag(Blasts) (p = 0.0002), PB-blast(%)(p = 0.0002) and IPF count (p = 0.0002) were independently associated with MDS. The area under curve (AUC) for RBC count (0.942), platelet count (0.881), NESSC (0.876), and NEFSC (0.865) showed higher diagnostic power with acceptable sensitivity and specificity in receiver operation characteristic curves (ROC) analysis. When using combinations of ‘NESSC and NEFSC and RBC count and platelet count, the positive predictive value (PPV) markedly increased to 100.0% (100%), however, the sensitivity decreased to 50.8%. Overall, if combination criteria satisfying one of four CPD cut-off criteria were used, all 64 MDS patients could be detected (sensitivity 100%). [Conclusion] This study demonstrates that CPD representing cell volume (decreased NEFSC) and complexity of neutrophils (decreased NESSC) may be useful for screening MDS using PB. The existing known markers, low RBC count and platelet count showed highest diagnostic efficiency in screening MDS. The combination of these four markers markedly improved sensitivity of MDS detection. These parameters are convenient and objective markers obtained from the automated analyzer with the CBC test. This could help to suspect the presence of MDS rapidly without additional cost.

¹School of Medical Laboratory Science, Health Institute, Jimma University Jimma, Ethiopia, ²Jimma Medical Center, Jimma University Jimma, Ethiopia

Introduction: The international consensus group suggested criteria for action following automated complete blood count and white blood cell differential analysis. These criteria were set based on data from laboratories of developed countries. It is highly important to validate the criteria in developing countries where infectious diseases are still rampant that can affect blood cell count and morphology. Thus, this study aimed to validate the consensus group criteria for slide review at Jimma Medical Center, Ethiopia from November 1, 2020 to February 30, 2021. Method: The study comprised a total of 1685 patient samples from the daily laboratory workload of CBC analysis. The samples were collected in K2-EDTA tubes (Becton Dickinson) and analyzed using Coulter DxH 800 and Sysmex XT-1880 hematology analyzers. A slide review was done on two Wright-stained slides for each sample. All statistical analyses were performed using SPSS version 20 software. Result: There were 39.8% positive findings, the majority of which were related to red blood cells. The false negative and false positive rates for Sysmex and Coulter analyzer were 2.4% versus 4.8%; and 4.6% versus 4.7%, respectively. The false negative rate was unacceptably higher when we used physicians' triggered slide review, which was 17.3% and 17.9% for Sysmex and Coulter analyzers, respectively. Conclusion: Generally, the consensus group rules are suitable to use in our setting. However, we might still need to modify the rules, particularly to reduce the review rates. It is also necessary to confirm the rules with case mixes proportionally derived from the source population.

¹Department of Clinical Genetics, Leiden University Medical Center Leiden, Netherlands, ²Labor Staber, Medizinisches Versorgungszentrum für Humangenetik Regensburg, Germany

Introduction:–After the introduction of MLPA as a standard tool in molecular diagnostics for thalassemia, a variety of cases emerged to have duplications of the complete α-gene cluster, including the conserved regulatory element HS-40. Increased α-globin synthesis leading to chain imbalance can have an aggravating effect in carriers of β-thalassemia variants by converting a typically asymptomatic carrier state into a thalassemia intermedia. Methods and Results–Here we describe a 58.9 kb segmental duplication of the entire α-globin gene cluster in a patient with low/borderline hemoglobin (10.4–11.7) while having normal ferritin and iron parameters. This finding was found using standard MLPA and was confirmed with and investigated further using CNV analysis from PCR free whole genome sequencing data. Breakpoint analysis shows a tandemly oriented local duplication, but with a probable inversion of only the duplicated HBZ gene in the duplicated gene cluster, thus suggesting a more complex rearrangement. Interestingly, HS-40 is not involved in the duplication, in contrast to most segmental duplication cases known from literature. Additional breakpoint PCR and Sanger sequencing will follow to confirm these findings. Conclusion–Segmental duplications of the entire α-globin gene in general have no effect on the hematology in carriers, as is the case in our patient. It remains unclear how this duplication affects the α-gene expression as the HS-40 is quite remote from the α-genes in the duplicated segment. One explanation could be that HS-40 divides its function over the duplicated clusters, resulting in equally distributed α-gene expression. Further studies need to be done to understand how the expression of the duplicated two α-genes is regulated.

¹Erythrocyte Diagnostic Laboratory, Cancer and Blood Diseases Institute, Cincinnati Children's Hospital Medical Center Cincinnati, OH, USA, ²Division of Hematology, Cincinnati Children's Hospital Medical Center Cincinnati, OH, USA

Introduction: Hereditary hemolytic anemias include red blood cell (RBC) membrane disorders (including hydration defects), RBC enzymopathies, and hemoglobin (Hb) disorders, due to pathogenic variants in genes coding for RBC membrane proteins, enzymes, or globins, respectively. Osmotic gradient ektacytometry (OGE) evaluates the deformability of RBCs as they are subjected to defined shear stress in a medium of increasing osmolality, in a laser-diffraction viscometer, and is a reference method for differential diagnosis of the erythrocyte membrane disorders, when a recent transfusion does not interfere with the phenotypic evaluation of the patient. Three distinct features in the OGE curve are: Omin, which corresponds to the value of the hypotonic osmolality where 50% of the cells hemolyze in an osmotic fragility assay and provides information on the initial surface to volume ratio of the cell sample; EImax, the maximum ElongationIndex that depends mostly on the cytoskeleton mechanics; and Ohyp, which is the osmolality value in the declining portion of the curve, where the cells' average maximum diameter is half of EImax, and reflects the initial intracellular viscosity of the red cells (Fig. 1). Methods: During the clinical use of OGE for patients evaluated for hemolytic anemias, using the LoRRca MaxSis Osmoscan (R&R Mechatronics International B.V.®, Zwaag, The Netherlands), we realized that the carrier status of a Hb disorder may affect the characteristics of the curve. Therefore, we conducted a systematic evaluation of this effect in samples of asymptomatic healthy subjects with heterozygosity for HbS (n = 15) or HbC (n = 7) in comparison to 15 healthy volunteer controls with no RBC disorder. Results: HbC trait causes a significant left shift of the ektacytometry curve with decreased Omin and Ohyp (Fig. 1), compatible with the dehydrated status of the erythrocytes containing HbC, and a slightly decreased EImax resembling curves diagnostic of hereditary xerocytosis. Blood samples of individuals with HbS trait had a normal ektacytometry in comparison to concurrently tested normal controls and the reference range of our laboratory. Conclusion: Hb disorders, even in carrier (trait) status, may affect OGE results. Comprehensive evaluation of the erythrocytes with complete blood count, reticulocyte count, smear review and Hb electrophoresis is important to avoid misinterpretation.

Russian Research Institute of Hematology and Transfusiology Saint-Petersburg, Russia

Introduction:. Thrombotic complications represent one of the main causes of morbidity and mortality in patients with Ph-negative myeloproliferative neoplasms (MPN) including essential thrombocythemia (ET), polycythemia vera (PV) and primary myelofibrosis (PMF). The pathogenesis of thrombosis in MPN is multifaceted anyway resulting in coagulation activation, platelets playing a great role in this process. The aim of the study was to characterize platelet functional activity in patients with Ph-negative MPN without treatment. Materials and methods. The study included 22 patients with Ph-negative MPN (5 ET, 6 PV and 11 PMF) without treatment with any medications at least 10 days before enrollment. Control group included 30 practically healthy volunteers. Platelet functional activity evaluation was performed in platelet-rich plasma using light-transmission aggregation induced by ADP (end concentrations 1 and 5 μM), collagen (2 μg/mL) or ristocetin (1.2 mg/mL). The aggregation initial velocity (V, %/min) and maximal amplitude (MA, %) were assessed. Statistical data processing was performed using STATISTICA 6.0. The results were presented as box-and-whiskers graphs with median, interquartile interval, minimum and maximum. Mann-Whitney U-test was used to compare patient groups with controls, Kruskal-Wallis H-test was used to patient groups comparison,the differences were considered statistically significant at p < 0.05. Results. Comparing with controls all patient groups had significant decrease in platelet aggregation induced by low ÐDP doses (1 μÐœ) (Fig. 1). The effects of other aggregation inducers including high ADP concentration differed in patients groups. The most pronounced changes characterized parameters of ET patients, velovity of aggregation induced by ADP (5 μÐœ) and collagen and MA of aggregation induced by collagen and ristocetin significantly exceeding control values. Comparing with healthy persons parameters of PV patients tended to increase, including velocity of aggregation induced by ADP (5 μÐœ) (p = 0.329) and МРof aggregation induced by collagen and ristocetin (p = 0.329 and p = 0.161 accordingly). In contrast to ET and PV, group of PMF patients was characterized by significant deceleration of aggregation induced by collagen comparing with either controls or ET patients (p < 0.017). Comparing with ET patients also velocity of aggregation induced by ADP (5 μÐœ) (p = 0.004) and MA of aggregation induced by collagen (p = 0.003) were significantly decreased in PMF while comparing with healthy controls these parameters as well as MA of aggregation induced by ADP (5 μÐœ) tended to decrease (p = 0.126; p = 0.317 and p = 0.078 accordingly). Along with indicated changes, suggesting platelet functional activity decrease in PMF patients, they had increased MA of aggregation induced by ristocetin comparing with controls (p = 0.055). Conclusion. Significant decrease in parameters of aggregation induced by ADP in low concentration in all MPN patient groups without treatment can indicate deficiency of platelet receptor apparatus. Pronounced increase of platelet aggregation activity induced by ADP in high concentration and other applied agonists in ET patients and a tendency to likely changes in PV patients as well as increased aggregation induced by ristocetin in PMF patients can play role in high thrombotic risk in MPN.

¹Department of Hematology, PGIMER Chandigarh, India, ²Department of Pathology, PGIMER Chandigarh, India, ³Department of Internal Medicine, PGIMER Chandigarh, India, ⁴Department of Pediatrics, PGIMER Chandigarh, India

Introduction: Clot Waveform Analysis (CWA) is the extended study of the slope generated by an optical detection system during coagulation tests. Multiple quantitative and qualitative CWA parameters of activated partial thromboplastin time (aPTT) provides valuable information in addition to clotting time. The present study was planned to assess the aPTT- CWA among lupus anticoagulant (LA) positive samples and whether the aPTT-CWA can differentiate LA positive samples from hemophilia A non-inhibitors (HANI) and HA with inhibitor samples (HAWI) Material and Methods: A total 69 LA positive samples were the part of study and among these 52 had prolongation of aPTT. The aPTT-CWA of these 52 samples were compared with 36 HANI and 16 HAWI samples. The CWA parameters (Fig. 1) were noted manually on the ACL-TOP 500 analyses screen using mouse pointer. Result: The quantitative aPTT-CWA parameters of LA positive samples were compared with HANI samples. The following were statistically significant by the Mann-Whitney U-test: acceleration peak time [1A], velocity peak time [1B], time for the height of fibrin formation [1C] and height of acceleration/ second derivative [-] peak [2B]. These parameters were significantly low in LA positive samples. Among the qualitative parameters: shoulder in first derivative, the biphasic wave in second derivative [-] and the serrated wave in second derivative were seen more significantly in HANI samples (Chi-square test). The quantitative aPTT-CWA parameters of LA positive samples were also compared with HAWI samples. The statistically significant by the Mann-Whitney U-test were: acceleration peak time [1A], velocity peak time [1B], time for the height of fibrin formation [1C], height of acceleration/ second derivative [-] peak [2B], height of fibrin formation (2D), width of acceleration [-] [3B] and width of velocity [3C]. All these were significantly low in LA positive samples. However, none of the qualitative parameters showed any significant statistical difference between these two groups. Conclusion: The quantitative aPTT-CWA parameters can be helpful in differentiating LA positive sample from HANI and from HAWI. The qualitative parameters are more beneficial in unraveling LA positive sample from HANI and not from HAWI.

Center for International Health Research, Rhode Island Hospital, Brown University Medical School, Providence, RI, USA

Stanford University, USA

The World Health Organization (WHO) and International Consensus Classification (ICC) recently updated their classification systems for myeloid neoplasms, including acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS), as well as lymphoid neoplasms, including high grade B-cell lymphoma. These updates have significant implications for the diagnosis and management of these diseases, and molecular laboratory testing plays a crucial role in identifying specific genetic mutations that inform treatment decisions. In this talk, we will discuss the implications for laboratory testing and how to navigate the subtle differences between the two systems. We will review the specific genetic mutations that are associated with a subset of these diseases, including those that are now included in the updated classification systems, and discuss the latest laboratory techniques for detecting these mutations. We will also discuss the challenges associated with implementing these updates in clinical practice and the need for continued research to refine and improve molecular laboratory testing, particularly at the boundaries of AML and MDS. This talk will be of interest to clinicians, laboratory professionals, and researchers who are involved in the diagnosis and treatment of hematopoietic neoplasia.

¹New York Presbyterian Hospital, Weill Cornell Campus New York, NY, USA, ²Marist College Poughkeepsie, NY, USA, ³Cornell Medical College New York, NY, USA

Introduction: The genetic loss of all four alpha chains (homozygous a⁰thalassemia; genotype: –/–) results in a lack of alpha chain synthesis, such that no hemoglobin A, A₂ or F can be generated, and the clinical syndrome of Bart's hydrops fetalis. Bart's hydrops fetalis is seen most frequently in Southeast Asia, where alpha thalassemia is prevalent but has been described in other areas. Usually, it is incompatible with life, with death in utero or shortly after birth. Antenatal diagnosis offers the option for in-utero therapy, which can result in a healthy live birth. However, the method treatment impacts therapeutic response–both based on laboratory findings and clinical response. Methods/results: The patient, currently 27 years old, was diagnosed with alpha thalassemia major (--SEA/--SEA) at 12 weeks gestational age and subsequently underwent both intra-uterine stem cell transplant (father's CD34+ stemcells) and intra-uterine red cell transfusions. The patient was born at 35 weeks gestational age via C-section with an exchange transfusion on day 1. Bone marrow biopsy showed no engraftment of the paternal stem cells. Although his physical development was delayed, in part to due to growth hormone deficiency, his intellectual skills were unimpaired. After receiving straight transfusions for many years, in 2015 the patient was changed to exchange transfusions. However, he soon rejected exchange transfusions due to time constraints. He received single donor units at 1–3-week intervals. During this time his hemoglobin level approached 14 g/dL, hemoglobin H levels rose to 86%, and nRBC count increased to almost 400/100 WBC, all indicators of a hematopoietic response. The hypoxia due to Hemoglobin H's extremely high affinity for oxygen and the hemolytic effect of precipitated beta tetramers attached to the RBC membrane, contributed to this response. Approximately 6 months ago he resumed monthly exchange transfusions. His hemoglobin H levels are now <10%, and his last hemoglobin value is 11.5 g/dL with<10 nRBC/100 WBCseen on smear. Conclusion: The expected lifespan of a viable newborn with alpha thalassemia major no longer needs to be 2 weeks. With prenatal identification of parental carriers of alpha thalassemia trait, in utero diagnosis can lead to birth of a healthy infant whose life expectancy is now unknown.

Haematology Laboratory, Department of Pathology and Laboratory Medicine, KK Women's and Children's Hospital Singapore, Singapore

Introduction: Iron deficiency anaemia (IDA) and thalassemia trait are common causes of microcytosis and hypochromia in children. It has previously been reported that Reticulocyte haemoglobin content (RET-He) is a useful parameter in the diagnosis of IDA. In this study, we aim to explore if RET-He can be used to differentiate between the two conditions, as there is a significant prevalence of thalassaemia trait in our local paediatric population. Methods: Non-gender specific, de-identified data from patients aged 2 to 12 years old with samples sent to the KKH Laboratory for workup of anaemia or thalassemia screening from Jan to Dec 2022 were retrospectively reviewed. Samples were screened for HbH inclusion bodies, haemoglobin electrophoresis (Sebia Capillarys3 OCTA Analyzer) and serum ferritin (Abbott Alinity Analyzer). IDA was defined as serum ferritin <20 μg/L with negative HbH inclusion bodies and normal Hb electrophoresis. The cohort with thalassaemia trait had either positive HbH inclusion bodies (alpha thalassaemia trait) or HbA2 > 3.2% (beta thalassaemia trait) with serum ferritin > 20 μg/L. The RET-He values of these cases were obtained and analysed. Descriptive statistics and graphics were analysed using SPSS Version 29.0.0.0 (241). Results: We collected a total of 81 samples meeting our selection criteria, with 34 IDA patients and 47 with either alpha or beta thalassemia trait. For the IDA cohort, the median RET-He value was 20.6 ± 1.1 pg with a wider spread of values ranging from 10.5 to 32.3 pg (Fig. 1). For the cohort with thalassaemia trait, the median RET-He was 20.3 ± 0.2 pg, ranging from 17.1 to 22.6 pg (Fig. 1). When comparing the mean RET-He values of both cohorts (20.5 pg for IDA vs. 20.1 pg for thalassaemia trait) using the Independent samples T-test, there was no significant difference with a p value of 0.717. Conclusions RET-He does not appear to reliably differentiate between IDA and thalassaemia trait when used as a standalone parameter in our local paediatric population.

Department of Laboratory Medicine, Hallym University College of Medicine & Hallym University Sacred Heart Hospital Anyang, Korea

Introduction: BCR-ABL1 negative myeloproliferative neoplasms (MPN) are characterized by the presence of driver mutations of JAK2, CALR, and MPL genes, which constitute diagnostic criteria. Recently, Next generation sequencing (NGS) enables to detect multiple genetic abnormalities but the effect of co-occurring mutations other than JAK2, CALR, and MPL genes remain unclear. In this study, we investigated the occurrence of co-occurring gene mutations and tried to figure out the impact of these genetic abnormalities on MPN. Methods: We retrospectively reviewed the NGS data of BCR-ABL1 negative MPN cases newly diagnosed from January 2021 to December 2022 in a university hospital. NGS was performed using gene panel of 108 genes associated with myeloid neoplasm. Total number of MPN was 28 cases: 7 polycythemia vera (PV), 16 essential thrombocythemia (ET), and 5 primary myelofibrosis (PMF). Results: Among 28 MPN cases, 18 cases (64.3%) had JAK2 mutation and two cases (7.1%) had CALR mutation and eight cases (28.6%) were triple-negative (TN). Co-occurring mutations were detected in 50.0% (14/28) of all MPN. There was no difference in the prevalence of co-occurring mutations between JAK2 mutation group and TN group (55.6% vs. 50.0%, p = 0.066). No co-occurring mutations were found in two CARL mutation-positive cases. Co-occurring mutations were present in 28.6% (2/7) of PV, 50.0% (8/16) of ET, and 80% (4/5) of PMF. The frequent mutations were TET2 (17.9%), DNMT3A, and MLLT3 (14.3% each), and ASXL1, SRSF2, and ZRSRS (7.1% each) in all MPN. TET2 (22.2% vs. 10.0%), DNMT3A (16.7% vs. 10.0%), and ASXL1 (11.1% vs. 0%) mutations tended to be more frequent in JAK2 mutation-positive group than TN group. Four (80%) out of five cases having TET2 mutation were ET, and SRSF2 mutations were only found in PMF Conclusions: Co-occurring mutations were found in more than half of all MPN cases and their prevalence was high in PMF and low in PV. TET2 mutations were frequent and mostly found in JAK2 mutation-positive ET cases. Although limited by the small number of cases studied, our results provide insight into the patterns of co-occurring mutations in MPN. Accumulation of data on co-occurring mutations will further strengthen the role of genetic abnormalities in clinical decision.

Mayo Clinic Rochester, MN, USA

Introduction::Andexanet alfa is a reversal agent for direct factor Xa inhibitors (DFXaI) approved for use in patients with uncontrolled or life-threatening bleeding. Although monitoring the DFXaI level is not necessary when administering Andexanet alfa, some providers have attempted to measure the DFXaI level after administration to monitor the patient's anticoagulant status. Anti-Xa activity assays with DFXaI-specific calibrators are commercially available to measure the DFXaI level in a patient's plasma. However, the assay utilizes a high dilution of the patient's sample initiating dissociation of the andexanet alfa from the DFXaI causing an inaccurate overestimation of the true unbound DFXaI level. In this study, we developed and validated a modified anti-Xa assay to mitigate this problem by reducing the dilution of patient sample in the test set up. Methods: A chromogenic anti-Xa kit (HemosIL® Liquid Anti-Xa Kit, Werfen, MA, USA) using apixaban calibrators and controls (HYPHEN BioMed, France) was optimized on the ACL TOP 700 (Werfen) to reduce the final sample dilution from 1 in 27.25 (reference) to 1 in 2.66 (modified). Intra/inter-assay precision was performed (two runs in duplicate, each day for 10 days) on three apixaban concentrations. Method comparison samples (n = 37) were tested on the modified and reference anti-Xa assay. Linearity was performed on one commercial and three de-identified waste samples diluted to at least seven concentrations. Reference range verification was performed following CLSI guidelines (n = 20). Lower limit of quantification (LLOQ) was assayed over multiple days with diluent buffer, normal pooled plasma, and control material diluted to ~10 ng/mL. Interferences were assessed using samples spiked with lipid, argatroban, and unfractionated (UF) heparin. In addition, studies were performed on apixaban patient plasma samples spiked with varying concentrations of andexanet and a patient plasma sample two hours post andexanet infusion. Acceptance criteria included: intra/inter-assay precision (CV ≤15%), method comparison to reference method (r² ≥0.950, slope 1.00 ± 0.10 and average % difference <20%), linearity (r² ≥0.950 and slope 1.00 ± 0.10), and LLOQ (CV ≤20%) on two reagent lots. Results: Precision results recovered CVs ≤2.4% (Intra) and ≤5.0% (Inter) achieved at all levels (21, 107, and 367 ng/mL). Method comparison displayed an r² = 0.958, slope = 0.988, and avg% diff = 6.6%. Linearity studies averaged an r² = 0.999 and slope = 0.995; assayed levels 9.3–482.5 ng/mL. Reference range verified <10 ng/mL and analytical sensitivity studies demonstrated LLOQ = 10 ng/mL. UF heparin interfered with the assay in concentrations ≥0.125 U/mL. Testing of the andexanet spiked apixaban patient samples and post infusion patient accurately demonstrated the expected decrease (average 78% reduction) in the apixaban levels (Fig. 1). Conclusions: The modified chromogenic anti-Xa assay met pre-defined performance specifications and can be utilized to measure apixaban concentration and unbound DFXaI levels after administration of Andexanet alfa in plasma samples from patients.

Medicine Laboratory Department. Hospital Universitari Germans Trias i Pujol, Badalona, Spain

Introduction: Peripheral blood smear (PBS) review is a very useful tool for the diagnosis of several hematological diseases. Currently, slidemaker stainers analyzers prepare slides automatically and allow to increase the productivity optimizing hematology laboratory workflow. Some of the stains contain methanol to fix the sample, which is an environmental pollutant. MCDh (Micro Chromatic Detection for Hematology) is a new set of four staining reagents based in Romanowsky type stains and adds the main advantage that they replace methanol with ethanol. The aim of this study was to evaluate the MCDh reagents for the staining PBS for clinical laboratory practice. Methods: The study was divided in two parts: First: we selected 48 samples with hemogram (EDTAK3) from oncological/hematological patients. Two PBS were prepared from each sample: one according to May GrümWald-Giemsa (MGG) and other according to MCDh protocols. The percentage of different cell types correctly classified by digital microscopy was calculated for both stains. Second: we selected 65 samples: 21 hemograms without quantitative/qualitative alterations (normal), 13 bacterial infection/sepsis, 3 viral infection, 7 chronic lymphoproliferative syndromes (CLPS), 5 myeloproliferative neoplasms (MN), 9 myelodysplastic syndromes (MDS) and 7 acute leukemias(AL). Six PBS were performed from each sample: three according to MGG and three according to MCDh protocols. Slides were reviewed by three different cytology experts and the concordance of morphological alterations (dysplasia neutrophils, toxic granulation, reactive and atypical lymphocytes) was evaluated between MGG and MCDh. PBS were performed by SMS-DxH analyzer (Beckman Coulter) and slides were processed by the Cellavision DM9600. Kappa index was used to evaluate the agreement between MCDh and MGG stains. Statistical analysis was performed by MedCalc Statistical Software-v19.5.3. Results: MCDh stain showed optimal results for cell classification by digital microscopy (table 1) and presented 55% less artifacts than MGG stain (149 and 329 artifacts, respectively). MCDh showed a good diagnostic capacity to detect all morphological alterations (Tab. 1). Atypical limphocytes were observed in CLPS (100%) and dysplasia neutrophils in different hemopathies and severe infections (alteration segmentation). Conclusions The MCDh stain is a good option to incorporate into the routine clinical laboratory to blood cells review, and detect atypical cells. Moreover, MCDh has the advantage that their reagents are metanol-free and peripheral blood smears present fewer artifacts.

¹Department of Pathology, Brigham and Women's Hospital Boston, MA, USA, ²Department of Pathology, Northwestern University Chicago, IL, USA, ³Department of Pathology and Laboratory Medicine, Emory University Atlanta, GA, USA

Introduction: Current flow cytometric analysis of blood and bone marrow samples for the diagnosis of acute leukemias relies heavily on manual intervention in both the processing and analysis steps, introducing significant subjectivity into the resulting diagnosis and increasing diagnostic turn-around time. Additionally, concurrent molecular characterization of these samples via cytogenetics and targeted sequencing panels can take multiple days, thereby delaying patient diagnosis and treatment. Attention-based multi-instance learning models are machine learning models that can make accurate predictions and generate interpretable insights regarding the classification of a sample from multiple events/cells; while these models have been developed for anatomic pathology applications, they have yet to be applied to flow cytometry data. Methods: By utilizing 1820 flow cytometry samples from 2019 to 2022 at Brigham and Women's Hospital, we developed attention-based multi-instance machine learning models for automated diagnosis of acute leukemia, including differentiation of acute myeloid leukemia (AML) from B-lymphoblastic leukemia/lymphoma (B-ALL). Additionally, using concurrent cytogenetic and targeted sequencing data from 674 acute leukemia samples, machine learning models for prediction of molecular aberrancies from flow cytometry data were developed. Machine learning models were created using the TabNet deep learning architecture and VIME self-supervised training algorithm, which are state-of-the-art approaches towards machine learning from tabular data including flow cytometry data. Results: Attention-based multi-instance models demonstrated strong performance for the automated diagnosis of acute leukemia versus non-leukemia samples (AUROC 0.869), as well as the separation of AML from B-ALL samples (AUROC 0.971). These models also accurately predicted cytogenetic aberrancies among AML samples, including t(15;17);PML::RARA (AUROC 1.00), as well as mutations including NPM1 (AUROC 0.725). These models do not require any manual intervention including compensation or gating, and additionally provide quantitative scores for the relative importance of different flow cytometry events and markers for the diagnosis of a particular sample. These importance scores can be integrated into flow cytometry analysis software for visualization and interpretation by hematopathologists. Conclusions: In this study, we have demonstrated the capability of machine learning models to provide automated diagnoses of acute leukemia, as well as accurately predict cytogenetic and molecular aberrancies in blood and bone marrow samples using flow cytometry data. This automated workflow can significantly decrease diagnostic turn-around time and ultimately improve patient outcomes.

First Affiliated Hospital, Sun Yatsen University, Guangzhou, China

Background: Few studies about Artificial intelligence (AI) have focused on medical education. The aim of the study was to evaluate the effectiveness of learning blood cell morphology by learning on our AI-based online platform. Methods: This study is based on a mixed-methods sequential explanatory design and crossover design. Thirty-one third-year medical students were randomly divided into two groups. Group A learned blood cell morphology on the AI-based platform in their spare time for one week before microscopy-learning. Both groups had microscopy learning together. Group B learned on the platform for one week after microscopy learning. Pretests and posttests were performed before and after platform-learning and microscopy-learning in each group, respectively. Students were interviewed, and the records were coded and analyzed by NVivo 12.0. Results: For both groups, test scores increased significantly after online-platform learning. Before microscopy learning, the platform could help the students learn blood cell morphology effectively with a large amount of blood cell images and detailed descriptions. After microscopy learning, the platform could help the students observe the rare cells in the blood smears. Feasibility was the most mentioned advantage of the platform. The AI classification could inspire the students to compare the similarities and differences between cells and help the students understand the cells better. Students had positive perspectives on the online-learning platform. However, most students thought that microscopy learning was indispensable. Conclusion: The AI-based online platform could assist medical students in blood cell morphology learning. It could be an effective and beneficial complement to microscopy learning. Students had very positive perspectives on the AI-based online learning platform. It should be integrated into the course and curriculum to facilitate the students.

¹Shandong Cancer Hospital and Institute, Shandong First Medical University and Shandong Academy of Medical Sciences, Jinan, China, ²Qilu Hospital, Cheeloo College of Medicine, Shandong University, Jinan, China

Background: The quality of lifehas attracted more and more attention. Due to the high cost of medical treatment and the need for long-term treatment, aplastic anemia (AA) is a serious threat to human health and has a multifaceted impact on the quality of life of patients. Objectives: The purpose of this study was to evaluate health-related quality of life in acquired aplastic anemia patients in China and analyze the influencing factors of health-related quality of life. Methods: A cross-sectional case-control study was conducted. We investigated 108 patients with acquired aplastic anemia. A total of 200 healthy subjects with matched sex and age were selected as control group. The SF-36 scale was used to compare the quality of life in the case group and the control group. SPSS 21.0 was used for statistical analysis of the data. The correlation between the scores of each dimension and the clinical data and general sociological data of the patients was analyzed. Results: The difference on statistical significance presented in five of eight dimensions of SF-36 between the AA patients and the normal population(p < 0.05), including physical functioning (PF), role limitations due to physical problems (RP), social functioning (SF), energy/vitality (VT) and mental health (MH).The quality of life of patients with platelet count>30 × 10⁹/L and hemoglobin level >80 g/L at follow-up showed no significant difference compared with the control group, p > 0.05). There were significant differences in the choice of treatment measures in all dimensions except BP (body pain, BP) (p < 0.05).The bone marrow/peripheral hematopoietic stem cell transplantation group had the highest quality of life scores. LSD mean multiple comparison showed that there was no significant difference in quality of life scores of patients after ATG/ALG, bone marrow/peripheral hematopoietic stem cell transplantation and umbilical cord blood transfusion, which were all higher than those in the cyclosporine treatment group (p < 0.001). Age, anemia, thrombocytopenia, bleeding symptoms, transfusion dependency, fatigue and main treatment were the influencing factors of quality of life, which were closely correlated with all dimensions of quality of life except BP, p < 0.05%. Conclusions: Quality of life is an important reference index to reflect the survival status, select precise treatment strategies and evaluate efficacy of AA patients. This study provides scientific basis for improving AA patients' quality of life.

¹Department of Laboratory Medicine, West China Hospital, Sichuan University, Chengdu, China, ²College of Electrical engineering, Sichuan University, Chengdu, China

Acute promyelocytic leukemia(APL) is a distinct subtype of acute myeloid leukemia (AML) characterized by its rapidly progressive and fatal clinical course if untreated and curable if treated timely. Screening out suspicious APL cases promptlyis vital toinitiate the treatment and overcome early death. We innovatively developed a framework consisting of the ResNet-18, a convolutional neural network (CNN) architecture, aiming to quantitatively map a complete blood count (CBC) scattergram to a probability indicating the susceptibility of APL fast and robustly. The validation was based on a dataset including 320 scattergrams of the white blood cell differential (WDF) channel from 51 APL patients, 510 from 105 non-APL AML patients and 320 from 320 healthy controls. A 5-fold cross-validation was performed by a random 4:1 stratified split into the training and validation datasets in each fold. In the 5-fold cross-validation, an area under the curve and average precision both > 0.99 were achieved in each fold (Fig. 1). Furthermore, 95% (304/320) of the scattergrams from APL were correctly flagged by our CNN-based model, which outcompetes the CBC review rules recommended by the International Society of Laboratory Hematology (all p < 0.001). To the best of our knowledge, our CNN-based framework is the first to utilize WDF scattergram output from routine CBC analysis to map suspicious APL cases early with outstanding sensitivity, specificity and precision. We further raise a new CBC workflow incorporating this framework upstream of the morphological review, which would provide the earliest flag for APL.

¹Department of Pathology & Laboratory Medicine, Oregon Health & Science University Portland, OR, USA, ²DeepCyto LLC West Linn, OR, USA

Introduction: Manual differentials of bone marrow aspirates (BM) are the gold standard for hematopoietic cell enumeration and evaluation for hematopoietic neoplasms. However, this process can be time-consuming and may vary depending on the examiners' experience. Artificial intelligence (AI) has significantly improved image processing and classification accuracy in the last decade. Currently, AI technology has been applied to various medical image interpretation tasks. The objective of this study is to evaluate the correlations of AI-assisted hematopoietic cell differential enumeration counts on BM aspirate smear with manual pathologists' counts. Methods: One hundred BM aspirate slides were retrieved from the archived diagnostic cases in the pathology department from 2018 to 2022. The study set includes 31 normal marrows, 27 cases of acute myeloid leukemia, 16 cases of myelodysplastic syndromes, 7 cases of myeloproliferative neoplasms, 5 cases of acute lymphoblastic leukemia, 10 cases of lymphoid neoplasms, and 4 cases of plasma cell neoplasms. The DCS-1800 Blood Cell Morphology Analyzer (DeepCyto, West Linn, OR) was used to scan, classify, and analyze 500 nucleated hematopoietic cells using proprietary algorithm. Cells were classified into five major hematopoietic categories: blast, myeloid, erythroid, lymphocytes, and plasma cells. The manual differential was extracted from the final pathology reports. Pearson correlation coefficients were used to compare manual and AI-assisted cell type enumeration. Results: The AI assisted BM differential counts showed strong correlation with paired manual analysis (Fig 1). The blasts and myeloid differential counts showed the highest correlation (r > 0.96 and r > 0.93, respectively). More importantly, AI-assisted analysis in this study set did not miss any acute leukemia comparing with manual counts. Erythroid and lymphoid differential counts revealed moderate correlation (r > 0.86 and r > 0.88, respectively). The plasma cell differential had lowest correlation (r > 0.76). Conclusions Our study demonstrated that AI-assisted BM differential can accurately classify nucleated hematopoietic cells across a broad range of hematological disorders. AI-assisted BM morphological evaluation has the potential to enhance the standardization, accuracy, and efficiency of morphologic evaluation of BM aspirates.

The Ohio State University

Peking University People's Hospital, Beijing

Introduction: The ALL patients with CNS involvement who received all grafting experienced a higher cumulative incidence of relapse (CIR), CNS relapse and an inferior survival compared to those without. FCM is superior to conventional cytology (CC) in identifying leukemia cells of cerebrospinal fluid (CSF). However, whether an isolated flow cytometry (FCM)-positive CNS involvement before transplantation holds clinical significance is unknown. Methods: 1406 ALL patients in complete remission (CR) prior to allo-HSCT which were classified into isolated FCM positive CNS involvement group, conventional cytology positive CNS involvement group and negative CNS group were analyzed retrospectively. Results: 31 patients (2.2%) were isolated FCM positive CNS involvement, 43 patients (3.1%) were cytology positive CNS involvement and 1332 (94.7%) were negative CNS. Among the three groups, the 5-year CIHR was 42.3% (95%CI 24.9%–59.7%), 44.2% (95%CI 29.3%–59.0%) and 22.3% (95%CI 20.1%–24.5%), respectively (p < 0.001). The 5-year CIR was 42.3% (95%CI 24.9%–59.7%), 48.8% (95%CI 33.9%–63.8%) and 23.4% (95%CI 21.1%–25.6%), p < 0.001. The isolated FCM positive CNS involvement group or cytology positive CNS involvement group had significantly higher CIHR and CIR than negative CNS group. The 5-year CIHR of pre-HSCT CNS involvement was 43.4% (95%CI 32.1%–54.7%, p < 0.001), and the 5-year CIR was 46.3% (95%CI 34.7%–57.4%, p < 0.001). Compare to the negative CNS, the 5-year LFS rate of pre-HSCT CNS involvement was lower [39.1% (95%CI 27.9%–50.3%) versus 60.8% (95%CI 58.3%–63.3%), p < 0.001], and 5-year OS rate was inferior [48.6% (95%CI 37.2%–60.0% vs. 66.0%, 95%CI 63.5%–68.5%, p = 0.001]. Multivariate analysis indicated that T-ALL, CR2 and above stage at HSCT, pre-HSCT MRD positive, and pre-HSCT CNS involvement were independent risk factors associated with higher CIHR, CIR and inferior OS, LFS. A new scoring system was developed using these four variables: low-risk group was 0, intermediate-risk group was 1, high-risk group was 2 and extremely high-risk group was 3 to 4. The 5 years of CIR was 16.9%, 27.8%, 50.9%, 66.7%, 5 years of LFS was 67.6%, 56.9%, 31.0%, 13.3%, and 5 years of OS was 71.8%, 63.4%, 38.3%, 13.3%, respectively (p < 0.001). Conclusion: Our results confirmed the patients with isolated FCM positive CNS involvement had a higher risk of recurrence after transplantation. And cases of pre-HSCT CNS involvement had a higher relapse rate and poorer LFS and OS rate.

UT Southwestern Medical Center

Older patients are more prone to develop anemia or mild cytopenias. For decades, little was understood about the etiology of these cytopenias in the setting of normal lab evaluation for vitamins /minerals and occasionally having normal bone marrow biopsies. Gene panel testing using next generation sequencing has shed insight into the presence of abnormal bone marrow clones carrying myeloid mutations without meeting criteria for a diagnosis of a myeloid neoplasm. In these presentation, we will cover various entities that involve clonal hematopoiesis, clonal cytopenia of undetermined significance (CCUS) and the current available prognostic tools that may aid in risk stratification, in addition to treatment options for those patients.

¹Department of Pediatrics, Nara Medical University Nara Japan, ²The Course of Thrombosis and Hemostasis Molecular Pathology, Nara Medical University Nara Japan, ³SEKISUI MEDICAL CO., LTD. Tokyo Japan

Introduction; Emicizumab (Emi), a bispecific antibody for hemophilia (H)A treatment, mimics as a cofactor function of activated factor (F)VIII and improves hemostatic effects. In activated partial thromboplastin time (APTT) measurement, it is difficult to differentiate Emi-treated patients from other patients based on the clotting times. If it can be presented as an Emi-contained plasma sample in the routine APTT measurement, it should be very helpful in the early detection of the appearance of anti-Emi antibodies and in the emergent treatment of Emi-treated patients under unconsciousness condition. The purpose of this study was to differentiate Emi-containing samples using APTT-based clot waveform analysis. Methods; Coagpia APTT-N and CP3000^TM (SEKISUI MEDICAL CO., LTD.) were used for APTT reagents and measurements. Total 69 patients in the Emi group (15 with and 54 without FVIII inhibitor; APTT 21.2-37.9 sec) and total 30 patients in control group [normal (10), congenital HA (1), acquired HA (4), HB (4), von Willebrand disease (3), FV deficient (4), Lupus anticoagulant (1) and others (3) within the reference range of APTT (24-39 sec)] were used. We applied two parameters "corrected Vmax × VmaxT" (corrected Vmax means the maximum velocity when the maximum value of the reaction curve is regarded as 100, and VmaxT means the time to reach the Vmax) and “R/L ratio” (L shows the time from 38% height of Vmax to reach Vmax, R shows the time to drop from Vmax to 38% height of Vmax). Differentiation was performed after SDI transformation of the parameter values of the Emi group using the mean and standard deviation of the control group. Results; Maximum velocity (Vmax) decreased in the Emi group compared to the control group. The Emi and control groups could be differentiated with 100% sensitivity and specificity using the "corrected Vmax x Vmax T (cut-off value: 1.8)" and "R/L ratio (cut-off value: 2.7)", Conclusions; We successfully differentiated the presence of all Emi-contained samples by using APTT-based clot waveform analysis. This method is expected to enable detection of Emi-treated HA patients' samples quickly in routine APTT measurement.

¹hospital Meixoeiro Vigo, Spain, ²laboratori Clinic Territorial ICS-IAS Girona, Spain, ³Group Neoma Facultat Medicina UDG Girona, Spain

Background Glucose-6-phosphate dehydrogenase (G6PD) deficiency is one of the most common X- linked recessive hereditary disorders in the world, affecting more than 400 million people worldwide. The enzyme catalyzes the first step in the hexose monophosphate pathway of glucose metabolism. The decrease as well as the absence of the enzyme increase red blood cells vulnerability to oxidative stress caused by exposure to certain medications (like sulfonamides) or intake of fava beans. Case report Starting from the index case of a male with glucose 6-phosphate dehydrogenase deficiency, the rest of the descendants in the family have been studied. The offspring of the index case have been four females, all four with glucose-6-phosphate dehydrogenase values of 2 IU/g Hb (the exact results are shown in the table below). Of these two women, both had a child with complete glucose-6-phosphate dehydrogenase deficiency (with a value of 0 IU/g Hb). The reference values for glucose 6-phosphate dehydrogenase are 5.70–9.90 IU/g Hb. The family tree with the affected individuals is shown in the image (the green arrow indicates enzyme deficiency). Conclusion Early detection and prevention is the key strategy for successful management and control of G6PD deficiency. The G6PD deficient individuals can be highly variable in its clinical presentation. Most of them are asymptomatic (until they are exposed to oxidative stress) so they transmit pathogenic mutated gene to the next generation

¹Hospital of the University of Pennsylvania Philadelphia, PA, USA, ²Children's Hospital of Philadelphia Philadelphia, PA, USA

Introduction: At the Children's Hospital of Philadelphia, approximately 37% of automated CBC differentials reflex to a manual differential when the ISLH Consensus Guidelines are applied. ISLH guidelines were developed for adult patients, but certain disorders, such as small B-cell lymphoma, are vanishingly rare in pediatrics. Given the time and resources involved, a target under 25% manual differentials is desirable. Methods: We evaluated the frequency of flagging for monocytosis and lymphocytosis. Historical rule: If lymphocytes >7K, perform manual differential with 1 day window. If monocytes >3K, perform manual differential with 1 day window. Data was collected from the electronic medical record on the frequency of these flags, the total differential counts and manual differential counts performed for the four months surrounding the rule changes. Results: Lymphocytosis flag triggered 415 and 412 times in the months prior to the change. 80% were on children under 2 years of age. None of these children had lymphoid neoplasms. Therefore, we changed the rule from a threshold of 7 to 10 K with a 30 day lookback window. The flag was triggered 101 and 101 times in the months following the change. Monocytosis flag triggered 161 and 175 times in the months prior to the change. 50% were on children under 3 months of age. No children had clinically significant monocytosis. Therefore, we changed the rule from a threshold of 3 to 4 K with a 14 day lookback. The flag was triggered 38 and 31 times in the months following the change. After implementation of these rule changes, the frequency of lymphocytosis and monocytosis flags across all differentials performed decreased from 5.4% to 1.3%. The frequency of manual differentials did not change significantly (35.3%, 36.4% prior and 36.9%, 34% after the rule change). Conclusions: Pediatric hospitals serve a unique patient population which may warrant adjustment of the ISLH guidelines for manual differentials. Changes to individual flags will decrease the frequency at which they are triggered, however due to individual specimens triggering multiple flags, this may not decrease the frequency of manual differentials. Additional investigation is required to discover other rule changes that may decrease the time and resource burden of manual differentials without negatively impacting patient care.

¹Laboratory Medicine Department. Hospital Universitari Germans Trias i Pujol Badalona Spain, ²Hematology Department. Institut CatalÃs d’Oncologia. Hospital Universitari Germans Trias i Pujol Badalona Spain

Introduction: Cell blood count (CBC) is one of the most commonly ordered laboratory routine tests and includes leukocyte, erythrocyte and platelet count. Most analyzers use Coulter principle to provide leukocyte count (LC), according to cell size. However, impedance LC can be interfered providing wrong results that can affect patients' follow-up and management. Other technologies can be used to measure LC alternatively. The aim of the study was to evaluate LC when impedance interference, using different technologies (impedance, optical and flow cytometry) and to suggest alternatives. Methods: DxH900 (Beckman Coulter) combines Coulter principle to determine CBC by impedance and VCS technology in a flow cell for leukocyte differential (DIFF). The analyzer offers four different LC: UWBC (raw measure), WBC (measure corrected by internal algorithms), WBCDOP (LC obtained in DIFF channel) and WBCNOP (LC from Nucleated Red Blood Cells–NRBC–channel). Under normal conditions, UWBC and WBC are equal. Between July 2022 and January 2023, 45 EDTA-K3 samples showing discrepancy between UWBC and WBC were selected. We determined CD45 by flow cytometry (Cytoflex), obtaining 5 LC for each sample: UWBC, WBC, WBCDOP, WBCNOP and CD45. Taking CD45 count as reference, we calculated the difference for all LC and obtained mean, median, maximum and minimum values of the differences. Passing Bablok regression was calculated. Results: Samples analyzed had WBC between 0.00 and 55.40 × 10⁹/L. In 67% of samples, interference was caused by platelet clumps, 15% giant platelets, 7% NRBC and 11% unknown. LC measured in NRBC channel (WBCNOP) showed the best results compared to CD45 count: average difference −0.56 × 10⁹/L (−1.45%) and median −0.06 × 10⁹/L (−1.23%). Results are shown in Tab. 1. Moreover, WBCNOPand CD45 were statistically comparable methods [y = 0.579(−0.107 − 1.769) + 0.857(0.704 − 1.014); r = 0.921, p < 0.0001]. Conclusions Interference detection in laboratory is crucial to provide reliable results. LC is essential in patients' follow-up and professionals must be aware of possible interferences and solutions. In this sense, DXH900 offers different LC, being WBCNOP a good option. However, this parameter is considered by manufacturer for investigation use only and is not reportable. Manufacturers should perform clinical trials and evaluation studies to validate investigation parameters with potential clinical utilities.

¹Department of Pathology and Molecular Medicine, McMaster University Hamilton, ON, Canada, ²Hamilton Regional Laboratory Medicine Program Hamilton, ON, Canada, ³Division of Hematology-Oncology, Department of Pediatrics, CHU Sainte-Justine, Université de Montréal Montréal, QC, Canada, ⁴Department of Clinical Laboratory Medicine, OPTILAB Montréal-CHU Sainte-Justine, Université de Montréal Montréal, QC, Canada, ⁵Department of Medicine, McMaster University Hamilton, ON, Canada

Introduction: Quebec platelet disorder (QPD) is an autosomal dominant bleeding disorder with a platelet-dependent gain-of-function defect in fibrinolysis. The cause is a duplication mutation of PLAU, that repositions a downstream, megakaryocyte-specific enhancer element, resulting in a > 100-fold increase in PLAU expression during megakaryopoiesis. Many of the clinical and laboratory features of the disorder can now be attributed to this unique fibrinolytic defect. For example, the lower platelet counts in persons with QPD improve on antifibrinolytic therapy, and return to baseline about 10 days after therapy. A relationship between the fibrinolytic defect and the platelet aggregation abnormalities in QPD has not been established. Persons with QPD show reduced aggregation and/or an absent secondary aggregation wave with epinephrine, with or without additional aggregation defects (e.g., with ADP and/or collagen) that resemble the effects of alteplase and reteplase lytic therapy on human platelet function. We postulated that the impaired QPD platelet aggregation function is directly related to the QPD fibrinolytic defect and would normalize during prolonged antifibrinolytic therapy. Methods: A father and son with QPD (confirmed by molecular testing for the QPD PLAU duplication mutation) were retested for aggregation abnormalities during prolonged tranexamic acid therapy. Platelet aggregation records from studies done twelve to thirteen years earlier were reviewed for comparisons. The father was tested several years after stating combined tranexamic acid and apixaban to prevent recurrence of thromboembolic stroke with hemorrhagic evolution, complicating atrial fibrillation. His affected son was tested 10 days after restarting tranexamic acid treatment for a right hip hemarthrosis. Results: Initial light transmission platelet aggregation tests for both individuals showed maximal aggregation responses (MA) of 15%–16% (reference range: 9%–100%), and the characteristic, absent secondary wave with 6 μmol/L epinephrine. The son's first test also showed mildly reduced MA with 1.25 mg/mL ristocetin (68%; reference range: 75%–100%) without von Willebrand factor deficiency. During tranexamic acid treatment, both had a biphasic aggregation response to 6 μmol/L epinephrine that reached 86%–92% MA, and the son's aggregation with 1.25 mg/mL ristocetin was normal (MA: 85%). Conclusion: Antifibrinolytic therapy for QPD restores biphasic aggregation responses to epinephrine, illustrating that like QPD bleeding, wound healing problems, and lower platelet counts, QPD platelet aggregation findings are a direct consequence of the platelet/megakaryocyte gain-of-function defect in fibrinolysis that normalizes with antifibrinolytic therapy.

Russian Research Institute of Hematology and Transfusiology, Saint-Petersburg, Russia

Background: Now it is well known that coronavirus infection is associated with thrombosis complications. COVID-19 area companied by “cytokine storm”, cell activation and damage. It can lead to microparticles (MP) release. MP expresses tissue factor (TF) and procoagulant phospholipids on their surface that can enhance prothrombogenic potential. Patients after COVID-19 may have hypercoagulability and risk of thrombosis complication for a while. It is interesting to investigate the rolÐμof MP in sustenance of prothrombogenic changes after COVID-19. Aim: Definition of MP significance for thrombin generation in patients after COVID-19 Methods: The study included 34 patients after mild and 58 patients after severe COVID-19. Patients after mild COVID-19 did not receive any antithrombotic prophylaxis with anticoagulant and/or antiaggregant medicines. All patients after severe COVID-19 were taken anticoagulant prophylaxis by direct oral anticoagulants. Reference blood samples were obtained from 30 healthy people. Thrombin generation was measured in platelet-free plasma by Calibrated Automated Thrombinography (CAT) according to Hemker et al. Following triggers were used: PRP-reagent, containing rTF (1pM), MP-reagent, containing negatively charged phospholipids (4 micromol), with CTI (40 micromol/ml).Еendogenous thrombin potential were evaluated (ETP, nM*min). Data were described by non-parametric methods using the median (Me), Q1 – Q3 interval and Mann-Whitney U test (Statistica 12.0). p < 0.05 was considered significant. Results are presented in the table. PRP-reagent and MP-Reagent allow to estimate thrombin generation which depends on procoagulant phospholipids and TF, respectively. Significant differences were found between patients after mild COVID-19 with controls and patients after severe COVID-19. Conclusion: Patients after mild COVID-19 without any prophylaxes have amplification of thrombin generation and increasing of prothrombogenic potential because of TF and procoagulant phospholipids expressed on MP. Patients after severe COVID-19 don't have rising of coagulation activity of MP, probably because of anticoagulant prophylaxis. That may suggest about role of anticoagulants in limitation of cell activation after COVID-19.

Colorado Coagulation, Labcorp Englewood, CO, USA

Introduction::Emicizumab (Hemlibra®, Genentech/Roche) is a bi-specific monoclonal antibody that acts as a surrogate for Factor VIII (FVIII), bridging activated Factor IX (FIXa) and Factor X (FX) to activate FX (FXa). Emicizumab treatment adjusted to body-weight has a predictable pro-coagulant response: routine monitoring is usually not required. In the instance of an unexpected bleeding event, and because of the expanding use of emicizumab worldwide, there is growing interest in availability of an assay to measure emicizumab levels. The aim of this study was to validate a modified One-Stage FVIII activated partial thromboplastin time (aPTT) assay to measure emicizumab in HA plasma. Methods: Emicizumab measured on the BCS®XP (Siemens Healthineers) used Dade® Actin® FSL (Siemens Healthineers) using emicizumab as a calibrator. The calibrator is assigned by r² Diagnostics, Inc. versus a master calibrator tied to the manufacturers (Genentech) potency assignment. Calibrator dilution by the instrument generates a 7-point calibration curve (2.5–100 μg/mL). Samples were tested in duplicate at the standard 1:8 dilution or 1:20 in saline (0.9 g/L). Diluted samples were mixed with immuno-depleted FVIII deficient plasma and aPTT reagent, and incubated at 37°C for 3 min. With the addition of calcium chloride the clotting time was measured. Three quality control levels were used: 12.5, 25 and 75 μg/mL emicizumab. Intra- and inter-assay accuracy and precision were assessed at 80.0, 40.0, 10.0, and 5.0 μg/mL emicizumab. Carryover, reagent stability, dilution linearity, selectivity, interferences and sample stabilities were assessed. Results: Intra- and inter-assay accuracy and precision were within ±20% for all emicizumab concentration levels. Dilution linearity was demonstrated up to a 1:4 pre-dilution in FVIII deficient plasma, and allowed a reportable range of 5.0 to 360 μg/mL. Carryover was not observed when employing the intensive wash function on the analyzer. The aPTT reagent system was stable for 8 h on the instrument, and samples were stable up to 6 h on a cold block, 6 h on the instrument, and for up to three additional freeze-thaw cycles. Long term sample stability at −20 and −70°C was demonstrated at 1 month; testing is ongoing. Neither hemolysis, icterus (40 mg/dL) nor lipemia (1400 mg/dL) interfered with the assay. Selectivity was demonstrated in ten haemophilia A donor plasma at 5.0 and 80.0 μg/mL emicizumab. Conclusions: A modified aPTT assay using emicizumab-specific calibration was successfully validated using the automated BCS®XP platform. This assay is suitable for diagnostic laboratory measurement of emicizumab in citrated plasma.

¹Siloam Hospitals Lippo Village Banten, Indonesia, ²University of Pelita Harapan Banten, Indonesia, ³Siloam Hospitals Kelapa Dua Banten, Indonesia

Introduction: Dengue infection is a common cause of acute febrile illness in Indonesia and has a wide range of clinical presentations, from subclinical to fatal Dengue Shock Syndrome. Diagnosis of dengue infection can be established by serological test which detects antigen and/or antibody based on days of fever. Leukopenia and thrombocytopenia are frequent findings in dengue infection and can be laboratory markers of severe dengue infection in critical phase which usually occur on day 3 to day 7 of fever. High fluorescence lymphocyte count (HFLC) is a research parameter in Sysmex which correlates with the presence of blue plasma lymphocyte which is characteristic of dengue infectiosn. We aim to assess sensitivity and specificity of HFLC in patients with dengue infection based on serological tests and the correlation of HFLC with leukocyte and platelet count in dengue infection as markers of severe dengue infection. Methods: A cross sectional retrospective study was conducted during January – August 2022 at Siloam Hospitals Lippo Village, Indonesia. Full blood count and HFLC data from Sysmex XN-1000 were obtained for 302 samples which were serologically positive for non-structural 1 antigen (NS1) dengue (105 samples), positive dengue antibody (50 samples), positive dual antigen-antibody (47 samples), and seronegative samples as controls (150 samples). Statistical analysis of data was done by SPSS ver29. Results: On receiver operator characteristic (ROC) curve analysis, we find that cut off value of HFLC <0.35% has sensitivity 61.9% and specificity 54.7% for positive antigen group, cut off <3.2% has sensitivity 82% and specificity 98.7% for positive antibody group, and cut off <1.85% has sensitivity 83% and specificity 92% for positive dual antigen-antibody group. Further analysis revealed that there was significant negative correlation between HFLC and platelet count (Spearman's correlation coefficient −0.501, p < 0.001), but no significant correlation between HFLC and leukocyte count. Conclusions HFLC is a valuable tool in screening dengue infection especially in critical phase which is in conjunction with the time of dengue antibody detection. Furthermore HFLC can be used as additional data for monitoring the patient during the course of disease.

¹Department of Medicine, McMaster University Hamilton, ON; ²Department of Pathology and Molecular Medicine, McMaster University Hamilton, ON; ³Hamilton Regional Laboratory Medicine Program Hamilton, ON; ⁴Nordic Biomarker Umea, Sweden

Introduction: Patients on direct oral anticoagulants (DOACs) are unlikely to benefit from anticoagulant reversal strategies with drug levels below 50 ng/mL (ISTH, 2016); however, drug-specific anti-Xa levels are not routinely available in most clinical laboratories. The MRX PT DOAC (PT DOAC) is a Research Use Only assay that measures the functional effect of DOACs by the clot-time ratio (CTR), which is a ratio between a DOAC-sensitive prothrombin time (PT) and DOAC-insensitive PT. The manufacturer CTR reference interval is 0.98–1.38. The goal of our study was to evaluate the sensitivity and specificity of the PT DOAC assay to detect DOAC drug levels >50 ng/mL. Methods: We retrospectively selected 65 residual clinical samples with known apixaban, edoxaban and rivaroxaban drug levels for further evaluation using the PT DOAC assay (Tab. 1). Samples were anonymized before testing. Reagents and quality control (QC) were provided by Nordic Biomarker (Umea, Sweden). Test protocols were set up on ACLTOP 750 (Werfen, Bedford, MA), and the clot time ratio (CTR) was calculated using Microsoft Excel (Redmond, WA). We calculated the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for a CTR >1.38 to detect a DOAC drug level >50 ng/mL with corresponding 95% confidence intervals (CI). Results: Transference of manufacturer CTR was acceptable. QC performance was within manufacturer ranges. The sensitivity (95% CI) of the PT DOAC assay to detect DOAC drug levels >50 ng/mL ranged from 75% to 100%, and specificity ranged from 44% to 86%. A breakdown of test characteristics by DOAC is shown in Tab. 1. Conclusions: The PT DOAC assay shows promise at excluding the presence of clinically significant DOAC drug levels, especially for edoxaban and rivaroxaban. Further validation is needed to confirm our findings, given our study's retrospective design and small sample size.

¹Department of Medicine, McMaster University Hamilton, ON; ²Department of Pathology and Molecular Medicine, McMaster University Hamilton, ON; ³Hamilton Regional Laboratory Medicine Program Hamilton, ON; ⁴Nordic Biomarker Umea, Sweden

Introduction: Direct oral anticoagulants (DOACs) are often prescribed in clinical practice. While routine monitoring is not recommended, some clinical situations may require a drug level. As specific drug levels are not routinely available, we evaluated the MRX PT DOAC assay as an alternative approach to measuring the functional effect of DOACs. The test principle is based on two prothrombin time (PT) assays run simultaneously, one DOAC-sensitive, the other DOAC-insensitive, and the clot time ratio (CTR) calculated. Our study aimed to evaluate the performance and feasibility of offering the assay in clinical laboratories. Methods: All reagents, calibrators and quality control (QC) were provided by Nordic Biomarker and are currently labelled Research Use Only (Umea, Sweden). Test protocols for DOAC-sensitive and DOAC-insensitive PT assays were programmed into ACLTOP 750 (Werfen, Bedford, MA). The DOAC-sensitive assay requires a 1:3 sample dilution, while the DOAC-insensitive assay has a final sample dilution is 1:21. Microsoft Excel (Redmond, WA) was used to calculate the CTR [DOAC-sensitive/DOAC insensitive results]. Our evaluation was limited to calibration, QC samples and individual normal donor plasma samples. Results: A single lot number of the reagent and QC material were tested. Following manufacturer direction, replicates of calibrators A, B and C determined acceptable ISI and normal clotting time (NCT) for the PT-DOAC insensitive assay, and the DOAC Sensitivity Index (DSI) and NCT for the PT-DOAC sensitive assay. Three levels of QC (low, normal and high (n = 11 per level)) had CTRs within manufacturer ranges. Individual normal donor plasma samples (n = 20) supported manufacturer CTR reference interval transference. Conclusions: Overall, the MRX PT DOAC performed well on the ACLTOP 750. A limitation of our study was that a single lot number of reagent was evaluated, and thus lot-to-lot variations were not assessed. Further study using patient samples on DOACs, other anticoagulants and interferences, such as lupus anticoagulant, are required. Calculating the CTR offline when run on ACLTOP analyzers is feasible for clinical laboratories.

Aga Khan University Karachi, Pakistan

Introduction: Prolonged PT and APTT are often observed in patients having bleeding symptoms or during the pre-operative workup. To differentiate clotting factor deficiency from factor inhibitor or lupus anticoagulants, PT/APTT mixing studies are performed. There is no consensus in what is considered corrected vs. non-corrected studies. In this study, we compared different approaches of interpretation of mixing APTT studies and evaluate the added value of mixing studies in patients requiring factor assay. Methods: APTT-1:1mixing studies during Dec 2022 to Jan 2023 were analysed through various approaches for correction and compared with our own criteria defined as a clotting time below the upper reference interval (URI). The studies were performed on fully automated coagulation analyzer CS 2500 (Sysmex, Singapore) before and after 2-hour-incubation at 37°C using lupus-insensitive reagent. Results were correlated with the specific clotting factor deficiency and or inhibitors as well as with ‘parallelism'. Results: A total of 48 mixing tests were evaluable after excluding 6 tests that showed prolonged PT only. There were 36 males and 12 females with a median (IQR1-3) age of 6 (1–18) years. Thirty two patients had an isolated prolonged APTT while both PT and APTT were prolonged in 16 patients. Overall, mean(±SD) PT was 29.4 (±45.5) and APTT was 82.3 (±43.9) seconds. Immediate and post-incubation correction was observed in 46 of 48 samples (95.8%) and factor deficiencies identified were VIII (n = 25), V (n = 8), IX (n = 6), X (n = 5), XII (n = 2), XI and XII (n = 1), and fibrinogen (n − 1). Two patients with uncorrected APTT-mix were diagnosed with acquired haemophilia A and showed ‘non-parallelism'. Comparison of URI correction strategy in immediate and incubated tests respectively showed greatest discordance with approaches like ratio-comparison (24, 37%), 5-seconds difference (33,41%) as compared to Rosner index (RI) (6.5, 9%) and percentile correction (11, 13%). Increasing the cutoffs to 1.4, 10 seconds and 18 for APTT ratio, APTT-time-difference and RI respectively improves the sensitivity of these approaches in discriminating clotting factor deficiency from inhibitors. Conclusion: Correction based on URI, appropriately identified the clotting factor deficiency and the presence of inhibitors while Rosner Index showed the best correlation with our approach. Various techniques for interpretation of mixing studies gave variable results and can be improved by re-defining the cut-offs for correction by the laboratory.

Core Laboratory. Biochemistry and Molecular Genetics Department. Biomedical Diagnostic Center, Hospital Clinic of Barcelona, Barcelona, Spain

Introduction: The presence of inmature blast cells in peripheral blood (PB) is normally related with malignant hematological diseases. Current hematology instruments have the capability to detect inmature cells analyzing the leukocyte population. These flags allow to select potential pathological samples in which the PB film review is mandatory to confirm or discard the presence of blasts. OBJECTIVES To analyze the utility of blast flag in the new analyzer Mindray BC 6800 Plus and to compare its performance with the blasts flag provided by Advia 2120i. Methods: A total of 143 K₂EDTA samples were analyzed in both Mindray BC 6800 Plus and Advia 2120i instruments. Samples were grouped according to the patient underlying pathology: (1) Myelodysplastic-Myeloproliferative Syndromes (MDS-MPS), (2) Acute Leukemia (AL), (3) Infections with the presence of Reactive Lymphocytes (RL) circulating in blood and (4) Group control of healthy patients (CONTROL). Sensitivity and specificity of blast flag were calculated in both analyzers for every group of patients. The presence of blasts in the PB film review was considered as the ground truth. Results: Mindray BC 6800 Plus showed a global sensitivity and specificity values of 92.3% and 73.9%, respectively, while the Advia 2120i showed sensitivity of and specificity of 73.1% and 90.8%, respectively. Mindray BC 6800 Plus showed a higher sensitivity with respect to Advia 2120i in the presence of blasts, even when low percentages are circulating in PB as occurred in MDS/MPS group. However, blast flag specificity was higher in the Advia 2120i considering normal samples and those from patients with infections and some reactive lymphocytes circulating in PB. Fig. 1 shows the concordance between the analyzers blast flag and the manual blast counts for samples corresponding to the different groups included in the study. Conclusion: The Mindray BC 6800 Plus analyzer showed greater sensitivity for the detection of circulating blasts in blood, although they are found in small percentages. However, the specificity was higher on the Advia 2120i, so blast flags provided by this analyzer are highly associated with hematologic malignancy.

¹Department of Electronics, Pontificia Universidad Javeriana Bogotá, Columbia, ²CORE Laboratory, Biochemistry and Molecular Genetics Department, Biomedical Diagnostic Center, Hospital Clinic, Barcelona, Spain, ³Department of Mathematics, Technical University of Catalonia Barcelona, Spain, ⁴Ressolve S.A.S Bogotá, Columbia

Introduction: The use of generative image models, specifically denoising diffusion probabilistic models (DDPMs), has gained significant interest in several applications. The improvements in diffusion model research have made DDPMs highly valued for their ability to generate high-quality and diverse samples. The purpose of this work is to augment existing data and compare it across eight types of peripheral blood (PB) cells using diffusion modelling techniques. Methods: PB smears were stained with May Grünwald-Giemsa and cell images were captured using the CellaVision DM96 analyzer. The data set, collected by our group and publicly available [1], comprises a diverse range of cell images including neutrophils eosinophils, basophils, lymphocytes, monocytes, immature granulocytes (metamyelocytes, myelocytes and promyelocytes), erythroblasts (NRBC) and platelets, with a total of 12 000 images. To balance the number of images of each type of cell, we used a total of 1000 images for each class. This set is divided into 70% for training, 10% for validation, and 20% for testing. The study employed a Diffusion Probabilistic Model, consisting of an autoencoder and a diffusion transition network, to remove noise from the data. The autoencoder compressed cell images into a latent space, which was then decoded back into the image space. A U-Net convolutional neural network was used to denoise the latent space and the autoencoder was frozen. The model generated synthetic samples for each cell type and was evaluated using 1600 images from the test set. The results showed the model's ability to reconstruct cell images effectively. Results: Using the test set, we use the Peak Signal Nosie Ratio (PNSR) metric to evaluate image reconstruction quality. In medical imaging a PSNR value above 25 is considered good, while a value above 30 is considered excellent. Fig. 1 shows some examples of the artificial generation of the PB cell images. The table in Fig. 1 shows that, eosinophils, basophils immature granulocytes and NRBC present higher PSNR values, preserving the relevant information in the images and reducing the amount of noise. Conclusion: We demonstrate the efficacy of using diffusion models in generating synthetic images of blood cells These models can help to significantly improve the accuracy of artificially generated images and contribute to improving the training of automatic recognition systems, particularly in cases of low-prevalence hematological diseases.

References

[1] Acevedo A, et al. A dataset of microscopic peripheral blood cell images for development of automatic recognition systems. Data in Brief 30 (2020).

¹JSLH Standardization Sub-committee for Blood Cell Counting, Tokyo, Japan, ²Nagasaki University Hospital, Nagasaki, Japan, ³JSLH Standardization committee, Tokyo, Japan, ⁴Nagasakiken Medical Association, Nagasaki, Japan, ⁵Kansai University of Health Sciences, Osaka, Japan, ⁶Beckman Coulter, Tokyo, Japan, ⁷BD Biosciences, Tokyo, Japan, ⁸Keio University Hospital, Tokyo, Japan, ⁹Shiga University of Medical Science Hospital, Shiga, Japan, ¹⁰Kinki University Hospital, Osaka, Japan, ¹¹University Hospital Kyoto Prefectural University of Medicine, Kyoto, Japan, ¹²Osaka Medical College Hospital Osaka, Japan, ¹³Kyoto Prefectural University of Medicine, Kyoto, Japan, ¹⁴Aichi Medical University, Nagakute, Aichi, Japan, ¹⁵ Japanese Society for Laboratory Hematology, Tokyo, Japan, ¹⁶The University of Tokyo, Tokyo, Japan

Introduction: The internationally recognized flow cytometry reference method for leukocyte differentials (RM-Diff) is established and used. It is important for practical use to confirm the convergence of the assay values for fresh blood calibrators measured by different laboratories. The Japanese Society of Laboratory Hematology (JSLH) executed assay value measurement in 12 laboratories to provide the target range for external quality assurance survey in Nagasaki prefecture. Methods: Fresh whole blood samples were collected from 2 healthy volunteers within blood bag containing CPDA solution, and EDTA-2K was added at a final concentration of 1.5 mg/mL. Aliquot volume (1.5 mL) of the samples were refrigerated and hand-delivered to each laboratory. The target assay ranges were determined by means of assay values measured by 13 flow cytometes (FACS CantoII 3, FACS Lyric 4, FACS Symphony 1 and Navios 5). The mean of three immunostaining tubes in three peripheral blood samples were used. The RM used the JSLH method which was validated by the ICSH method. All examiners were trained using the RM procedure provided by JSLH. The samples were measured within 24 h after drawing. Target assay values were determined by means of assay values, and the target assay ranges as allowance interval (AI) was determined from %Bias. %Bias was calculated by Intra-individual biological variation (CVI) and between individual biological variation (CVG). The CVI and the CVG were used values in the biological variation database by the European federation of clinical chemistry and laboratory medicine (EFLM-BV). %Bias was calculated by 0.25 x (CVG² +CVI²)^1/2). The convergence of the target assay range as analytical variation (CVA) was compared with CVI. Results: The AIs of sample A for %neutrophils, %lymphocytes, %monocytes, %eosinophils and % basophils were 52.93 (49.29–56.56), 39.18 (36.72–41.64), 4.74 (4.43–5.04), 1.74 (1.45–20.3) and 0.86 (0.80–0.92). The AIs of sample B for them were 62.97 (58.65–67.29), 28.53 (26.73–30.32), 5.45 (5.09–5.80), 1.27 (1.05–1.48) and 1.31 (1.21–1.40). All CVA were less than 0.5 CVI. Conclusion: The convergence of the assay values for fresh blood calibrators measured by different laboratories were within precision requirement of EFLM criteria of “desirable”. The AIs was determined by the mean of assay values and %Bias calculated CVI and CVG in the EFLM database.

Department of Hematology, Hospital Universitari Vall d’Hebron, Vall d’Hebron. Instituto de OncologÃa (VHIO), Vall d’Hebron Barcelona Hospital Campus, Barcelona Barcelona, Spain

Introduction: Monocytosis is a frequent finding in hematology labs that often requires additional slide review under the optical microscope (OM). Owing to the great majority of monocytosis have a reactive origin an early and reliable method to identify neoplastic cases is of great interest, particularly for the diagnosis of chronic myelomonocytic leukemia (CMML). The MWO application is a new solution available in the Sysmex XN software aimed for the automated differentiation between reactive and neoplastic monocytosis. Aim: The goal of the present study was to evaluate the MWO application and to compare its results with OM and flow cytometry (FC) for the diagnosis of CMML. Methods: A MWO score was automatically calculated in the analyzers of XN9000 hematology line. A score <0.160 (−) considers the monocytosis as reactive, being automatically validated. By contrast, a score ≥ 0.160 (+) is suggestive of neoplastic monocytosis, therefore generating a blood smear for OM review. Two cohorts of samples were established: the first, used to validate the threshold score of MWO, was reviewed by OM for the presence or absence of dysplasia (OM+ and OM−, respectively). The second cohort was simultaneously studied by OM, MWO, and by flow cytometry (FC). Results: We studied 104 patients (65 males, median age 70 year [25–97] Among them, 77 were MWO− (76 OM− and 1 OM+) and 27 MWO+ (16 OM+ and 11 OM−). Using the cutoff of 0.160, MWO had a sensitivity and specificity of 94.1 and 87.4% respectively, negative predictive value 98.7%, and an accuracy of 88.5% for the detection of pathological (dysplastic?) monocytosis with an optimal cut-off of 0.16. The second cohort comprised 26 patients from the Hematology Day Care Unit. Correlation for monocyte count between the three methods (MWO, OM, and FC) was excellent: ICC = 0.8 (95%CI: 0.69–0.289). Five patients had an established diagnosis of CMML. In 4 of them (80%) MWO, OM and FC was concordant. In the remaining 21 patients, 12 (57%) were MWO+ of whom 7 (33%) had a FC suggesting CMML. Of note, median of MWO was higher in patients with CMML (MWO ≥ 0.787 vs. 0.201). Conclusions: MWO is an accurate method to differentiate pathologic from reactive monocytosis, potentially increasing the laboratory efficiency by reducing the number of smear reviews. Using a threshold >0.160, MWO was concordant with FC in 80% of cases with CMML. Larger studies are ongoing that will ascertain the utility of MWO in identifying pathological monocytosis.

University of Utah/ARUP Laboratories

This talk will focus on practical considerations in bringing machine learning to the clinical flow cytometry laboratory including development, validation, and maintenance of pipelines.

¹University of British Columbia Vancouver, BC, Canada, ²Department of Pathology, Saint Paul's Hospital Vancouver, BC, Canada

Introduction: Acquired hemophilia A (AHA) is a rare bleeding disorder caused by autoantibodies (inhibitors) directed against coagulation factor VIII (FVIII) [1]. The incidence is approximately 1 per million [2, 3]. Spontaneous bleeding is common and may be severe and life threatening [5], underscoring importance of prompt diagnosis. Because of its rarity, there is limited data pertaining to laboratory diagnosis of AHA. Methods: Saint Paul's Hospital (SPH) is the provincial referral center for special coagulation testing in British Columbia, Canada. Data was collected retrospectively from patient medical records. Diagnosis of AHA was made using a combination of clinical history, an isolated prolonged activated partial thromboplastin time (aPTT), aPTT mixing studies, decreased FVIII activity (one stage clot based method) and inhibitor detection using Bethesda method [6]. Results: Over a 5 year period (2018 and 2022) our laboratory received samples from 23 patients with an isolated prolonged aPTT that were diagnosed with AHA. Similar numbers of men and women were observed (48% female) with a median age of 76. At initial presentation, median hemoglobin was 98 g/L and median aPTT was 70 s (range 48 to 108 s) where aPTT was ≤50 s for 3 (13%) cases. Within 24 h of reporting first prolonged aPTT, interpretive comments for prolonged aPTT plus additional investigations (mixing study +/- Bethesda assay) were performed in 52% of cases, interpretative comments for prolonged aPTT only were issued in 26% cases and no interpretative comments, mixing studies, or other investigations were performed in 22% of cases. Ultimately, aPTT mixing study was performed in 91% of cases including a 1 h incubated mix in 81% of cases. Immediate mix showed no correction in 52% of cases, partial correction within 6–8 s of control in 38% of cases and correction within 5 s of control in 10% of cases. All cases showed some degree of prolongation after 1 h incubation. Mean prolongation was 21 s (range: 5 to 44 s) where 24% of cases showed prolongation <10 s. Median inhibitor and factor VIII levels were 43 Bethesda units (range: 1 to 1525) and 0.04% (range: <0.01 to 0.18), respectively. Median PTT at the time of Bethesda assay was 74 s (range 50 to >140). Five (22%) cases were originally identified during lupus anticoagulant testing. Conclusions: Laboratory testing is the cornerstone for diagnosis of AHA, yet timely identification remains a challenge. Laboratory data from this cases series highlights opportunities for improvement that we are hoping to implement provincially.

¹Hematology Department, Sao Paulo State University UNESP Botucatu, Brazil, ²HEMACORE–Hematology Center Sao Jose dos Campos, Brazil, ³Pathology Department, AC Camargo Cancer Center Sao Paulo, Brazil, ⁴Patology Department, Sao Paulo State University UNESP Botucatu, Brazil

Bone marrow (BM) reaction to various stimuli can involve both parenchymatous and stromal cells due to the complementary interplay between these two components of hematopoietic tissue. Therefore, the laboratory presentation of peripheral blood (PB) smear and BM may show considerable variation in morphological presentation, depending on the various etiologies involved. We present PB and BM morphology from 12 patients (4 children and eight adults) whose parenchymal-stromal alterations were related to: (1) COVID-associated hemophagocytic syndrome; (2) parvovirus B19; (3) AIDS and miliary tuberculosis; (4) AIDS and visceral leishmaniosis; (5) miliary Leprae; (6) disseminated, juvenile Paracoccidiodomycosis brasiliensis; (7) folate deficiency associated with: (i) Down syndrome, (ii) inflammatory intestinal disease, (iii) and the use of antiepileptic drugs; (8) cytopenias associated with: (i) the use of azathioprine after kidney transplantation, (ii) and methimazole for Graves' disease; and (9) inflammatory syndrome (MDS x VEXAS?). We describe the clinical-hematological (parenchymatous) alterations due to stromal impositions whose cytohistological evaluation diagnoses the etiologies of cytoses and/or cytopenias, when the exclusion of primary hemopathies, such as acute leukemias, myelodysplastic syndromes, and other BM insufficiencies are imperative. This study illustrates some damaged bone marrow patterns, mainly due to non-clonal stimulus, in cases referred to a general hospital, recorded at the Laboratory of Hematological Morphology, of the Department of Hematology and Hemotherapy, UNESP, a public university, where our routine laboratory exams are performed.

¹Hirosaki University Graduate School of Health Sciences HIROSAKI, Japan, ²Hirosaki University Hospital HIROSAKI, Japan, ³Hirosaki University Graduate School of Medicine HIROSAKI, Japan

Introduction: Lymphocyte morphological changes can be reactive or neoplastic. Reactive morphological changes are atypical lymphocytes, which are lymphocytes activated and juvenile by antigenic stimuli from external enemies such as viral infection. Although it is important to distinguish normal lymphocytes from atypical lymphocytes in selecting clinical treatment, morphological differentiation is often difficult, even for experts. The aim of this study is to develop a new technology that can differentiate atypical lymphocytes from normal lymphocytes with high accuracy. We studied the possibility of an AI model for blood morphology classification using a deep learning method. Methods: The ResNet-101 model was applied to deep learning. The original image set for AI training consisted of 6393 typical nucleated blood cell images. Nucleated blood cell images were classified into nine categories (Band/ Segment/ Eosinophil/ Basophil/ Monocyte/ Normal Lymphocyte/ Atypical Lymphocyte/ Abnormal lymphocyte/ Erythroblast). The data augmentation process was applied to the original images, and transfer learning and fine-tuning were performed on the AI model. The subjects for clinical assessment were 20 cases of healthy persons and 20 cases of patients with atypical lymphocyte appearance. The Cut off value for cases in which atypical lymphocytes appeared was set at 3% or higher. We evaluated the agreement between visual classification and AI classification. Results: The clinical assessment results of healthy cases were shown in Tab. 1A, and patient cases with atypical lymphocyte appearance were shown in Tab. 1B. The total accuracy for healthy cases was 0.937 and 0.780 for patient cases. The Recall and Precision for atypical lymphocytes in this AI model were 0.620 and 0.857, respectively. Both sensitivity and specificity were 100% for differential diagnosis between healthy persons and patients with atypical lymphocytes. Conclusion: This AI model is highly accurate in differentiating cases of atypical lymphocytes and is considered extremely useful for peripheral blood screenings in clinical laboratories. However, the AI misrecognized 38% of atypical lymphocytes, most of which were classified into the normal lymphocyte category. Characteristics of misrecognized cells were borderline-type cells with moderate or gradient cytoplasmic basophilia. Our future work is to improve the accuracy of atypical lymphocyte recognition. We will consider other deep learning techniques or new categorizations suitable for AI classification in differentiating borderline atypical lymphocytes.

¹Hirosaki University Graduate School of Health Sciences Hirosaki, Japan, ²Hirosaki University Hospital Hirosaki, Japan, ³Hirosaki University Graduate School of Medicine Hirosaki, Japan

Introduction: Lymphocyte morphological changes can be reactive or neoplastic. Neoplastic morphological changes are observed in the peripheral blood as abnormal lymphocytes or lymphoblasts, which have a morphology similar to immature cells. Acute lymphocytic leukemia (ALL)/lymphoblastic lymphoma progresses rapidly. Therefore, it is essential to detect abnormal lymphocytes in blood screening tests and start medical treatment immediately. However, abnormal lymphocytes showing various morphological changes may be difficult to differentiate. This study aims to develop a new technology that can differentiate abnormal lymphocytes from normal/reactive lymphocytes with high accuracy. We studied the possibility of an AI model for blood morphology classification using deep learning. Methods: ResNet-101 model, a convolutional neural network-based object detection algorithm, was used to generate AI models for blood cell classification. The original image set for AI training consisted of 6393 typical nucleated blood cell images. Nucleated blood cell images were classified into nine categories (Band/ Segment/ Eosinophil/ Basophil/ Monocyte/ Normal Lymphocyte/ Atypical Lymphocyte/ Abnormal lymphocyte/ Erythroblast). The data augmentation process was applied to the original images, and transfer learning and fine-tuning were performed on the AI model. The subjects for clinical assessment were 15 cases of ALL patients. The Cut off value for cases in which abnormal lymphocytes appeared was set at 3% or higher. We evaluated the agreement between visual classification and AI classification. Results: The clinical assessment results of ALL patient cases are shown in Tab. 1. Total accuracy, Ave.F-measures, Ave.Precision and Ave.Recall was 0.825, 0.685, 0.702, and 0.801, respectively. This AI model's recall and Precision for abnormal lymphocytes were 0.977 and 0.812, respectively. Most of the 292 normal lymphocytes that the AI model misidentified as abnormal lymphocytes were small lymphocytes with high N/C ratios and strongly basophilic cytoplasm. Both sensitivity and specificity were 100% for differential diagnosis between healthy persons and patients with abnormal lymphocytes. Conclusion: This AI model was highly accurate in differentiating cases of abnormal lymphocytes and was considered extremely useful for peripheral blood screenings in clinical laboratories. However, this AI model had issues differentiating normal small lymphocytes from abnormal small lymphocytes. We consider that this issue requires hybridizing the monoclonality characteristic of tumor cells into the AI decision-making process. We will continue to improve this AI model and apply it to screening techniques for lymphocytic neoplasms such as chronic lymphocytic leukemia and malignant lymphoma.

University of Utah/ARUP

Our understanding of lymphomas has deepened exponentially due to rapid advancements in molecular phenotyping. In particular, the accessibility and reproducibility of genomic profiling by next generation sequencing has allowed for hundreds of studies which support newer molecular classifications and prognostication systems for common and rare lymphoid neoplasms. Keeping up to date on relevant molecular advancements is challenging, but feasible. Here we will focus on specific areas which significantly overlap with practical current and near future application in lymphomas.

¹U.O.C. Laboratorio Analisi GOM REGGIO DI CALABRIA, Italy, ²Instituto Clinico "Prof. De Blasi" Reggio Di Calabria, Italy

Introduction: Persistent polyclonal B-cell lymphocytosis (PPBL) is a rare lymphoproliferative disorder, usually benign, characterized by chronic polyclonal lymphocytosis of B cells. This syndrome as usually reported in young female smokers although it has also observed in isolate males and in a non-smoker female with absolute lymphocytosis on blood counts. The syndrome is generally without symptoms, although the presence of splenomegaly and hepatomegaly has been reported. It consists of a moderate, absolute lymphocytosis exceeding 4 × 10⁹/L in the peripheral blood for more than 6 months due supported by polyclonal expansion of mature B cells, and is accompanied by the presence of IgM type polyclonal hypergammaglobulinemia, and expression of the haplotype HLA-DRB1*7. The morphological picture of PPBL is characterized by the presence of element of medium size and a variable quota of binucleated lymphocytes in peripheral blood. Several studies have shown that the presence of binucleated cells and atypical lymphocytes in the peripheral blood was highly suggestive of a viral infection, such as with the Epstein-Barr virus. In this report we have described a case of PPBL with IgM EBV+. Methods and results Because of a suspected lymphoproliferative disorder, the sample of peripheral venous blood (SVP) of a 34-years old male with leukocytosis and persistent lymphocytosis arrived at the laboratory of Flow-Cytometry of the G.O.M. Reggio Calabria (Italy). The blood count showed leukocytosis (18 × 10⁹/L, lymphocitosis (72.6%) and mild thrombocytopenia (140 × 10⁹/L). Fluorescence analysis was performed by the flow-cytometer BD FACSCanto II with eight colors and two physical parameters with the lise wash method and the data have been processed using the FACSDiva software. Flow-Cytometric examination showed an increase of the B lymphocyte population CD19+ (Fig. 1) CD20+ (Fig. 2)CD5- (Fig. 1) polyclonal for Ig κ/λlight chains (Fig. 3) (as reported in hierarchy Fig. 4), plausibly associated with PPBL. Morphological examination of the SVP streak showed some medium-large lymphocytes with bilobed nucleus typical of PPBL (Fig. 5) confirming this diagnosis. Subsequent investigations confirmed increase in policlonal IgM and positivity of IgM EBV+. Conclusions In the diagnosis of PPBL,the laboratory of Flow-Cytometry and the observation of cell morphology, plays a role of fundamental importance to accompany the clinician in the formulation of the correctly diagnosis to avoid unnecessary therapeutics initiatives and prevent inappropriate diagnosis of malignant lymphoproliferative disorders.

Aga Khan Hospital, Kisumu, Kisumu, Kenya

Introduction: Transfusion of blood and blood products can be a life-saving intervention. Numerous medical interventions rely on reliable supply of adequate and safe blood products. The World Health Organization estimates that over 118.54 million units of blood are collected in over 13 000 blood collection centres in 169 countries each year ¹. Though useful and life-saving, transfusion of blood and blood products carries inherent risks of transfusion reactions and transmission of infectious agents. Over 68 bacteria, viruses, protozoa and prions can be transmitted from donors to recipients if appropriate precautions are not observed ². Effective donor selection criteria, sensitive screening tests, optimal processing and pathogen inactivation needs to be employed to ensure safe use of blood and blood products³. Methods: This is a 8-year restrospective single institution audit of donor data from 1 January 2014 to 31 December 2021 at Aga Khan Hospital in Kisumu, Kenya. Donor characteristics (age, gender, blood groups, rhesus factor and type of donor) and rate of seroprevalence of four transfusion-transmitted infections over the 8 year period were evaluated. Data was analyzed and presented in form of charts and graphs. Results: The most common type of blood donors were voluntary donors (57.9%), family replacement donors (40.5%) with low numbers of apheresis and autologous transfusions at 1.5% and 0.1% respectively. Blood group O was most common blood group at 58%, followed by A, B and AB at 20%, 18% and 3% respectively. Rhesus negative rate was at 3%. Hepatitis B was the most common detected infection at 2.5%, followed by HIV at 1.3% among all screened units. There were very low numbers of hepatitis C virus, malaria, and syphilis. There was a total of 402 (4.2%) infected blood units among the 7187 units transfused for the period evaluated. All transfusion reactions were reported as minor and classified mostly as febrile non-haemolytic and allergic reactions. No major reactions were reported. The average rate of transfusion reactions was 0.7%, with a range of 0.5%–1%. Conclusion: There was a high rate of transfusion transmitted infections (TTIs) in our setting with hepatitis B and HIV being the two most common TTIs. Continued improvement donor selection, better testing technologies and continued increased hemovigilance is needed in blood banking to ensure high quality and safe blood products.

Queen's University, Canada

Inherited bleeding disorders are diagnosed primarily on the basis of hemostasis lab testing together with a bleeding phenotype which can vary in severity. However, a significant minority of patients with bleeding symptoms will remain without a definite diagnosis even after extensive hemostatic testing. These are referred to as bleeding disorders of unknown cause (BDUC). Molecular genetic testing aids diagnosis as it can confirm a phenotypic diagnosis, distinguish between disorders with similar bleeding phenotypes and provide important information in particular situations such as prenatal diagnosis and detection of the carrier states. Genome-wide approaches help identify novel genetic associations that may explain the cause of bleeding in the patients without established diagnoses. This talk will highlight the scope of genetic disorders of coagulation and the responsible genes, the purpose of genetic testing for coagulation disorders and available methods, the role of molecular genetic testing in common and rare disorders and finally the challenges with genetic testing for coagulation disorders and important considerations.

Manipal Hospital Bangalore

Public health challenges including the SARS-CoV-2 pandemic have been instrumental in preparing low and middle income countries to set up molecular laboratories especially for PCR based assays.

These countries also carry a huge burden of hematological diseases like Chronic Myeloid Leukemia. There are various programs which offer free tyrosine kinase inhibitors to CML patients in the developing world, however, free or subsidised testing is not easliy available for evaluation of molecular response.

Despite financial and infrastructural constraints, resource limited countries must gear up to provide testing services for monitoring response to therapy for CML patients. This presentation shares our experience in setting up a PCR laboratory with bare minimum and at the same time continue to maintain all aspects of good laboratory practise.

¹Department of Laboratory Medicine, Dong-A University College of Medicine Busan, Korea, ²Department of Neurology, Dong-A University College of Medicine Busan, Korea

Introduction: Direct oral anticoagulants (DOAC) medication in atrial fibrillation (AF) decreases risk of the development of ischemic stroke. However, larger thrombi might form under insufficient anticoagulation during DOAC therapy. The aim of the present study was to investigate whether plasma DOAC concentration at admission correlate with the presence of large cerebral artery occlusion (LAO) in acute ischemic stroke patients with AF taking DOAC. Methods: We enrolled all patients admitted with acute cerebral ischemia while taking rivaroxaban, apixaban or edoxaban for AF between March 2021 and May 2022. The site of arterial occlusion was evaluated by initial magnetic resonance angiography. Plasma levels of DOAC were measured using the calibrated Xa-activity and categorized into low (<100 ng/mL), or high (≥100 ng/mL). We analyzed plasma DOAC concentration in relation to occurrence of LAO using correlation and multivariable regression analyses. Results: Of the 38 patients with ischemic stroke on DOAC (median age = 76 years, interquartile range IQR = 69–79; median plasma DOAC concentration = 92.8, IQR = 29–171.2), acute LAO on admission was diagnosed in 14 patients and plasma DOAC levels were found to be low in 19 (50%) and high in 19 patients (50.0%). LAO were associated with lower plasma DOAC levels (low: 29 ng/mL, IQR = 29–93.3 vs. high: 162.8 ng/mL, IQR = 88.9–195.6; p = 0.001). Low DOAC plasma levels were an independent predictor of LAO (odds ratio, 7.33 [95% confidence interval (CI), 1.58–33.9]; p = 0.011). Plasma DOAC levels were associated with prothrombin time (Spearman rho = 0.548, p < 0.001) and activated partial thromboplastin time (Spearman rho = 0.489, p = 0.002). Conclusions: Lower DOAC plasma concentration was associated with increased risk of LAO in acute ischemic stroke patients with non-valvular atrial fibrillation taking DOAC.

¹Department of Laboratory Medicine, Dong-A University College of Medicine Busan, South Korea, ²Department of Pediatrics, Dong-A University College of Medicine Busan, South Korea

Introduction: Current practice for laboratory investigation of hemoglobinopathies include automated complete blood count (CBC) and high-performance liquid chromatography (HPLC) or capillary electrophoresis (CE) for Hb (Hemoglobin)A₂ and HbF measurement and detection of Hb variants. Capillary 3 OCTA(Sebia) is one of the popular tools for identifying hemoglobin variants. For neonatal cases, a dedicated reagent is required. However, several Institutions are conducting newborn hemoglobinopathies screening using Capillary 3 OCTA without the dedicated reagent, and there is no consensus for reference interval. The purposes of this study are to test and validate the Capillary 3 OCTA on EDTA samples and to establish a reference interval for neonatal HbF by Capillary 3 OCTA. Methods: Capillary 3 OCTA was evaluated for precision, carry-over and comparison. Also, 23 neonatal samples were separated and measured with Capillary 3 OCTA. The Capillary 3 OCTA requires HbA₂ for measurement of HbA and HbF. Therefore, the newborn samples lacking HbA₂ were progressed to an additional test, mixed with the normal control sample in a 1:1 ratio to provide HbA₂. The reference interval of HbF was calculated by the value that processed without normal control. Results: With respect to the precision of Capillary 3 OCTA, the test results for low- and high- concentration control showed a coefficient of variation of less than 5%. HbF results for Capillary 3 OCTA and HPLC system Tosoh G11 showed a good correlation (R² = 0.9984). There is a limit that reference interval for neonatal HbF has not been established in the Capillary 3 OCTA. The results demonstrated that the HbF mean value of 83.2% and SD 8.1%. The 95% confidence interval was 80.9% to 85.6%. Conclusion: The Capillary 3 OCTA demonstrated reliable analytical performance with good precision, carry-over and comparison results. In this study, we identified the Capillary 3 OCTA is reliable for screening test for hemoglobinopathy in newborn, especially useful in institutions that already had Capillary 3 OCTA for hemoglobinopathy.

¹Center for Innovation in Diagnostics, Siemens Healthcare Pvt Ltd, Bengaluru, India Bengaluru, India, ²Dr. Jariwala Laboratory & Diagnostics, Mumbai, India Mumbai, India, ³Siemens Healthcare Products GmbH, Marburg Germany Marburg, Germany, ⁴Siemens Healthcare Pvt Ltd, Mumbai, India Mumbai, India

Introduction: The objective of this study was to determine the correlation between key parameters obtained from two different instruments: Atellica® HEMA 580 Analyzer and the ADVIA® 2120i Hematology System. Both instruments use flow cytometry but vary in their measurement technology. The Atellica® HEMA 580 uses impedance, white light absorbance and fluorescence detection for cell classification and enumeration, while the ADVIA® 2120i uses laser light scatter and myeloperoxidase activity. Methods: Hematology parameters were obtained from the Atellica® HEMA 580 and ADVIA® 2120i instruments for 203 samples from a single center, Jariwala Labs, in Mumbai, India. Statistical analysis to compare measurements from the two instruments was performed using Spearman's rank correlation and Passing-Bablok regression¹. Results: Spearman's rank correlation of >= 0.93 was obtained for WBC, RBC, MCV, hemoglobin, platelets, neutrophils (%), lymphocytes (%) and eosinophils (%); 0.84 for monocytes (%) and 0.75 for basophils (%). The slopes for all parameters except monocytes (%) were close to 1, indicating no significant differences in measurement values between the two hematology systems. The slope for monocytes (%) was found to be 1.64, which could be due to the differences in measurement principles and cluster analysis for WBC differential. Out of 203 total samples, 7 had WBC <4 × 10³ cells/μL and 30 had WBC > 10 × 10³ cells/μL. Due to small sample size, separate analysis of abnormal WBC samples was not performed. Conclusions: This single-center study demonstrates strong concordance between CBC and WBC differential results between the two analyzers.

Brigham and Women's Hospital, USA

Digitalization has become vital for many, if not all, clinical pathology operations spanning from a specimen collection to rendering a diagnosis. However, fully digital clinical pathology workflows are few and far between. This presentation will review the advantages and limitations of the current state of digital clinical hematology, with a specific focus on the novel technologies and AI-based algorithms that will enable full digitalization of peripheral blood smears. Digitalizaiton today leads to digital transformation tomorrow!

Mayo Clinic, Rochester

Dr William L. Nichols, MD was a hematologist at Mayo Clinic in Minnesota. His expertise was in clinical and laboratory evaluation and management of coagulation disorders particularly platelet disorders and von Willebrand disease. He had a national and international recognition as an expert in coagulation disorders and a unique ability to assimilate research findings into clinical practice and education; HIs ability to teach and simultaneously provide skilled clinical care was exemplary, he had a visionary leadership as chair of the Coagulation Diseases Group at Mayo Clinic and was very creative development of coagulation laboratory test panels to enhance care of patients with coagulation disorders. His firm commitment to the team based approach to patient care and his respect for each member of the team led to his being leading the International Hemophilia treatment centers (WFH) Mayo Clinic director. He mentored many younger colleagues in career growth. HIs academic accomplishments included leading the NHLBI sponsored effort on Diagnosis and Management of Von Willebrand disease. He is survived by his wife Sharon and children and grandchildren.

Cleveland Clinic

Several known obstacles in accessing medical data are directly tied to strict but understandably required regulatory restrictions within this space. HIPAA in the US and GDPR in the EU are in place to protect such patient privacy concerns and help prevent the abuse of sensitive data. Although these regulations are necessary, they also make it difficult for innovators and researchers to acquire and use such data in a timely manner to progress their quality studies and healthcare research.

Major advancements in machine learning have greatly moved forward the advent of synthetic data creation which will hopefully help overcome some of the aforementioned obstacles. The use of such synthetic data can drastically expedite certain research and quality studies within our healthcare arena. Synthetic data should not be confused with deidentified real data. As oppose to the deidentified real data that can be theoretically reidentified, synthetic data is entirely comprised of “new” data that does not represent any specific individual patient but rather their collective mathematical and statistical relationships. The final objective is a synthetic data that is able to mathematically and statistically approximate its real data counterpart and to test an idea for a given pilot study. The use of such data will not only dramatically expedite data access to the individual investigators but will markedly lower the entry barriers in healthcare research and innovation.

University of Utah School of Medicine, ARUP Laboratories

Hereditary hemolytic anemia (HHA) is a heterogeneous group of disorders due to genetically caused defects in red blood cell membrane structure, enzymes, heme and globin synthesis, erythroid proliferation, and differentiation. Traditionally, the diagnostic process has been complex and included many tests from routine to highly specialized ones. A significant phenotypic overlap between different HHA types introduces even more complexity. The inclusion of molecular testing has significantly improved the diagnostic yield. As more molecular modalities become available for clinical use, it is imperative to understand their benefits and disadvantages pertaining to the HHA diagnostics. Re-evaluation of the traditional diagnostic workflow may also bring forth additional benefits. This presentation will focus on the current state of molecular testing for HHA and its place in the diagnostic workup.

¹Laboratori Clinic Territorial ICS-IAS Girona, Spain, ²Group Neoma, Facultat De Medicina Girona, Spain, ³Institut Catal〠Oncologia Girona, Spain

Introduction: Iron deficiency anemia (IDA) and anemia of chronic diseases (ACD) are the two most prevalent anemia worldwide. On ACD, there is functional iron deficiency (ID) due to a key regulator, hepcidin. The presence of IDA in ACD patients can cause deleterious symptoms and must be supplemented. Since biochemical parameters are affected by inflammation, there is the need to find new indices to asses IDA in these patients. New erythrocyte indices are not influenced by this factor and are useful to diagnose ID status and monitor iron supplementation. Methods: We performed a cross-sectional study studying 124 anemic subjects. We studied differences among these anemic groups using Mann-Whitney and Kruskall-Wallis tests, considering a p < 0.05 to be statistically significant. Receiver operating characteristic analysis (ROC) curves and area under the curve (AUC) was calculated to obtain specific cut-offs for these parameters. Results: For Ret-He we propose a cut-off value of 30 pg to differentiate between ACD and ACD + IDA, with sensitivity of 88% and specificity of 75%, and an AUC of 0.899. For %Hypo He, we considered a cut-off value of 0, 35% with specificity of 72% and specificity of 75%, and an AUC of 0.798. For %MicroR we obtained a cut-off at 1.25% with specificity of 75% and specificity of 72%, and an AUC of 0.816. Discussion Ret-He showed the best performance and strong agreement was observed with other publications. For %Hypo He and %MicroR, we observed differences on cut- off values as we are studying ACD+IDA instead of just IDA or iron restricted erythropoiesis.

¹National Medical Research Center of Cardiology named after academician E.I. Chazov Moscow, Russia, ²Department of Biomedical and Molecular Sciences, School of Medicine, Queen's University Kingston, ON, Canada, ³School of Baccalaureate Nursing, St Lawrence College Kingston, ON, Canada, ⁴Clinical Pathology Department, Faculty of Medicine, Mansoura University Mansoura, Egypt

Background: Intercellular junctions (tight and gap junctions) are multiprotein complexes that provide contact or adhesion between neighboring cells or between a cell and the extracellular matrix. Coxsackie-adenovirus receptor (CAR), a protein of tight junctions, is expressed in various cells including cardiomyocytes and platelets. Connexins 43 and 45 are gap junction proteins. Studies using CAR knockout mice revealed that there is a cross talk between tight and gap junctions with altered expression and localization of Ñonnexins 43 and 45 in CARpr KO cardiomyocytes.In the myocardium CAR can moderate expression of connexins, proteins of gap junctions which were recently found in platelets.We aimed to study the interaction between CAR and connexins in platelets. Materials and methods: Eighteen patients (mean age 43.5 years, 20 male) with dilated cardiomyopathy (DCM) and 15 healthy subjects (HS) were recruited for the study. Platelet aggregation (spontaneous and ADP-induced) was studied by simultaneous analysis of the mean aggregate size changes and light transmission using an aggregation analyzer (230LA aggregometer). Platelet CAR, connexin 43 and 45 were assessed by immunofluorescence using anti-CAR, anti-connexin-43, and anti-connexin 45 antibody. Human myocardium was used as a positive control for connexins.The number of connexin positive platelets was determined relative to the total number of platelets and expressed as a percentage. Results: In all HS platelet aggregation was within normal whereas in majority of DCM patients there wasincreased spontaneous aggregation (83.3%). Low dose ADP-induced aggregation (0.1 μM and 1.0 μM), was seen in 15 (83.3%) and 14 (77.7%) patients respectively. The expression of connexin 43 was a rare event both in HS and DCM groups. Connexin 45 was observed in singular platelets and in platelets aggregates where it co-localized with CAR at sites of intercellular communication, and in leukocytes. Expression of connexin 45 on platelets was 7[5.5; 10.0]% in HS compared to 20[8.5;29.8]% in DCM group (p = 0.0001). In both groups platelets revealed filopodia with substantial amount of connexin 45 but weak staining for CAR. Conclusion: We provide evidence for novel expression of gap junction protein connexin 45 in normal and hyper-aggregable platelets as well as platelets' filopodia. Connexin 45 colocalizes with CAR at sites of intercellular communications forming contact membrane between neighboring cells. These finding indicate a possible role in cell signaling.

Purdue University

Flow cytometry has been a critical tool in hematology for well over 40 years. In those 40 years, the hardware changed little from year to year, mostly in incremental steps. This established a long term very stable technology, that was produced by several companies creating an environment where manufacturers competed, but all produced quality instruments. In the world of clinical flow cytometry however, there was a slightly different pattern. Only a limited subset of manufacturers created clinical diagnostic instruments and they also matched these instruments with appropriate reagents. With the development of spectral cytometry however, the core clinical manufacturers have not yet produced spectral instruments. One spectral instrument manufacturer has gained approval for use in China and Europe, but it is unclear how this will develop in terms of an instrument-reagent combination. With spectral flow cytometry expanding rapidly, this opens up the question of what impact significant technology changes might have on clinical cytometry.

Clinical Biochemistry Sevilla, Spain

Background: Perioperative coagulopathy and postoperative bleeding are the most common complications in patients undergoing cardiac surgery, especially when the cardiovascular surgery is associated with cardiopulmonary bypass (CPB). In this context, some studies suggest that implementation of viscoelastic point-of-care tests (POCT), such as rotational thromboelastometry (ROTEM, in conjunction with a specific algorithm for coagulation management, allow for better control of hemostatic pathology. Methods: Retrospective cohort study including 675 patients who underwent cardiac surgery with cardiopulmonary bypass. The incidence of clinical postoperative complications were analyzed before and after ROTEM® implementation. Results: Following viscoelastic testing and the implementation of a specific algorithm for coagulation management, the incidence of any allogeneic blood transfusion decreased (41.4% vs. 31.9%, p = 0.026) during the perioperative and postoperative period (26.5% vs. 19.2%, p = 0.061). In addition, significant reductions were detected in the incidence of heart disease (57.7% vs. 55.8%, p = 0.275; statistically significant reductions were detected in the incidence of postoperative pericarditis (3.6% vs. 1.2%, p = 0.043), postoperative renal failure (1.6% vs. 3.2%, p = 0.435), postoperative sepsis (1.2% vs. 0.9%), p = 0.337) and postoperative hematologic complications (postoperative bleeding (9.5% vs. 5.3%, p = 0.037), surgical reexploration (6.0% vs. 2.9%, p = 0.035).and length of Intensive Care Unit (ICU) stay (6.0 days vs. 5.3 days, p = 0.026). Finally, we observed a statistically significant decrease in the lengths of Intensive Care Unit (ICU) stay (6.0 days vs. 5.1 days, p = 0.026), after implementation of the POCT system and the specific algorithm for coagulation management. There were no statistically significant group differences with respect to total hospital stay (16.7 days vs. 13.5 days, p = 0.076). In-hospital mortality associated with cardiac surgery also did not change (4.5% vs. 2.4%, p = 0.132). Conclusion: The monitoring of hemostasis by ROTEM® in cardiac surgery, was associated with decreased incidence of allogeneic blood transfusion, clinical postoperative complications and lengths of hospital and ICU stay.

¹Abbott Santa Clara, CA, USA, ²University Medical Center Utrecht Netherlands

Director of Scientific and Medical Affairs, Abbott

Erythrocyte Diagnostic Laboratory, Cincinnati Children’s Hospital Medical Center Cincinnati, OH, USA

The ACKR1 gene encodes the atypical chemokine receptor-1, which is enriched within endothelial junctions and facilitates the transendothelial migration of neutrophils. Homozygosity for the single nucleotide polymorphism (SNP)-67T>C (rs2814778) in the promoter of the ACKR1 gene, also called the Duffy-null genotype, protects against Plasmodium vivax malaria. The Duffy-null genotype is also associated with lower circulating white blood cell (WBC) and neutrophil counts without predisposition to infections. This human variation, historically known as “benign ethnic neutropenia” is most common in people of African and Yemenite-Jewish ancestry, but it also occurs in other ancestral backgrounds. It is now referred to as the Duffy-null associated neutrophil count (DANC). We aimed to validate a clinical test for the identification of the Duffy-null genotype and evaluate its utility for the definitive identification of DANC in individuals referred for evaluation of neutropenia. The ACKR1 rs2814778 genotypes were discriminated as C/C, T/C, or T/T SNPs by real-time TaqMan based PCR in genomic-DNA extracted from EDTA-collected blood samples. Commercially available samples (n = 67) with known ACKR1 sequence was used during the validation phase. Genotype, demographic and hematologic data was collected for the first year of clinical use of the assay, with the inclusion of samples from healthy adult controls (T/T, n = 30) and three groups of neutropenic patients (n = 115). In this clinical validation study we demonstrated 100% accuracy and excellent precision using genotyped samples. The frequency of each genotype, among neutropenic subjects tested, was 74% CC (Duffy-null), 20% TT, and 6% TC. The most frequent ethnogeographic background reported for each genotype was 80% European-Americans for TT, 96% African-Americans for CC and 57% European-Americans for TC. As expected, patients had lower neutrophil counts than healthy controls. Among patients, there was no difference in neutrophils counts across the CC, TT, and TC genotypes (1.6 vs. 1.8 vs. 2.2 cells/μL × 10³), indicating that neutrophil count alone cannot distinguish between pathologic neutropenia and normal human variation (DANC). We also found the CC genotype was associated with low WBC but normal lymphocyte, monocytes and eosinophil counts. This assay was developed and used in our institution for the genetic determination of DANC in children and adolescents referred to our clinical Hematology service because of neutropenia. After a year of clinical use, the results confirm its utility and reliability, when combined with an appropriate clinical evaluation, helping to avoid prolonged, costly, and unnecessary work-up and follow-up. Genotype-specific reference ranges for leukocyte and neutrophil counts should be considered.

ARUP Laboratories/University of Utah

Healthcare has become increasingly digital since the passage of the HITECH Act in 2009. As a result, there is a growing need to optimize healthcare IT to allow for the interoperable exchange of data. As a result, the Office of the National Coordinator for Health IT has implemented their Final Rule for the 21st Century Cures Act. This requires certified health IT systems to use modernized messaging standards for the safe and secure exchange of data within health information networks and also requires the use of terminology standards including LOINC, SNOMED, and UCUM for coding clinical and laboratory data. Given the critical importance of laboratory results in the delivery of healthcare, laboratorians must become familiar with these principles of interoperability. Their clinical laboratory expertise is needed to appropriately structure and code test results to safeguard against improper aggregation or misinterpretation by downstream users and systems.

University of Washington

UT Health San Antonio

The increasing use of anticoagulants other than traditional vitamin K antagonists has led to a growing interest in the clinical utility of laboratory monitoring to enhance safety and effectiveness. Particularly, while direct oral anticoagulants such (DOACs) such as factor Xa inhibitors (FXaIs) do not have a regulatory requirement for laboratory monitoring, the use of FXaI-specific anti-Xa concentrations has gained traction and has been advocated for several indications. The laboratory and clinical service can encounter several challenges in the decision to use specific monitoring assays for alternative anticoagulants, in their performance, and in their interpretation. This presentation will focus on clinical scenarios in which monitoring of alternative anticoagulants may be informative and, on the inherent challenges.

West Medica Produktions- und Handels- GmbH Wiener Neudorf, Austria

Introduction: Diagnostic in hematology is based on increasingly specific techniques of morphology analysis. The correct interpretation of the blood smears viewed under a microscope is crucial. Hematologists typically focus on red and white blood cells, so digital image techniques are valuable aids in their analysis. The differential counting of the cells provides invaluable information to pathologists for the diagnosis of many diseases when the manual counting of the cells is time-consuming, tiresome, and susceptible to error approach due to the tedious nature of this process. Methods: The objective of the present study was to compare the results obtained from the automated whole slide imaging system with the Vision Hema Clinical Application Module to the reference results reported by anonymous pathologists. The target values and diagnoses were determined by averaging the test results of peripheral blood samples from the same subjects, reported by independent participants. The comparison was made between the counts (%) of white blood cells, including neutrophils, eosinophils, basophils, monocytes, lymphocytes, blasts, promyelocytes, myelocytes, metamyelocytes, and plasma cells obtained from the automated whole slide imaging system and the reference values. Sensitivity and specificity were used to describe the accuracy of the examination. Additionally, the differential and/or lead diagnoses given by hematologists using the automated whole slide imaging system were compared to the reference diagnoses. The interpretation of diagnostically significant findings included an assessment of red blood cells' size, color, forms, and inclusions. Four cases, including those with cancer and parasitological findings, were evaluated by hematologists using ×50 and ×100 oil-immersion whole slide imaging scans. Results: The determination of the selected white blood cells was in the target range except for the lymphocytes in the three of four cases. The lead and/or differential diagnoses were correctly established in all cases. Conclusions: We proposed automatic identification and pre-classification of white blood cells, and identification of red blood cells by size, color, forms, and inclusions. The automated whole slide imaging system demonstrates the efficacy of the proposed method for differential and/or lead diagnosis. The results were compared with reference results reported by an involved anonymous pathologist in the respective round. It's a useful tool that increases its adoption among hematologists, pathologists, scientists, educators, and research initiatives.

Division of Hematopathology, Department of Laboratory Medicine and Pathology, Mayo Clinic Rochester, MN, USA

Introduction: Hemoglobin (Hb) Hammersmith (Hb Chiba) is a rare hemoglobinopathy [B42(CD1)Phe-->Ser], HGVS HBB:c.128T>C, p.F43S predominantly reported in females. The altered protein tertiary structure favors deoxyhemoglobin Hb (T state) stabilization reducing oxygen affinity. This unstable Hb variant precipitates causing moderate to severe hemolytic anemia, splenomegaly, and transfusion dependence. Hb Hammersmith has been reported in published literature and disease-specific databases as electrophoretically silent. Contrary to this assertion, we have reproducibly detected data abnormalities that signal the presence of similar variants utilizing contemporary protein studies in our practice. In addition, correlation of methemoglobin and sulfhemoglobin levels provide further characterization. Methods: After IRB approval, a review of patient samples received for clinical hemoglobinopathy testing was undertaken. Protein studies included cation-exchange high performance liquid chromatography (HPLC), alkaline, acid and capillary electrophoresis (CE), heat and isopropanol stability studies, isoelectric focusing, mass spectrometry (MS) and methemoglobin/ sulfhemoglobin levels. The variant was confirmed using Sanger sequencing (HBB). Results: Four cases of Hb Hammersmith were identified. All were females with hemolytic anemia requiring transfusion. Two cases were related: a mother-daughter duo (Hgb 9 g/dL, MCV 104 fL) with pallor noted in the daughter at 6 months requiring splenectomy (age 4), decreasing but not abolishing transfusion need. The third case was a 30-month-old female (Hgb 7 g/dL) with history of cola-colored urine/jaundice without trigger, low pulse oximetry values, and no family history. The fourth case was a 16-month-old female (Hgb 8.6 g/dL) with persistent hemolysis. In all four cases, heat and isopropanol stability studies were abnormal. HPLC data were reproducibly abnormal (two small peaks at ≈4.27 and 5.10 minutes) although some variation in shape and elution time of the abnormal peaks was seen. CE showed a small peak in zone 1. IEF did not reliably detect the variant, but intact mass spectrometry showed a peak at mass 15 808 amu with approximate quantities 13%–19%. Hb F percentage varied (1.7% to 30%). Alkaline and acid gel electrophoresis were normal. Methemoglobin (2.2%–3.4%) and sulfhemoglobin (2.2%–3.6%) levels were elevated. Conclusions Hb Hammersmith displays reproducibly abnormal tracings by HPLC and CE and is detectable by intact MS. It is associated with low pulse oximetry readings and increased methemoglobin and sulfhemoglobin levels, supportive of an M-hemoglobin variant classification. Finally, although commonly de novo, we report a mother-daughter pair with autosomal dominant inheritance. Hb Hammersmith shows similar laboratory features as other unstable oxygen affinity-altering variants, such as Hb Köln, and can be detected by contemporary methods.

The Children's Hospital, University of Child Health Sciences Lahore, Pakistan

Introduction: Pediatric acute leukemia accounts for 25.1% of all cancers in children. Aberrant immunophenotypes is the pattern of antigen expression on neoplastic cells different from the process of normal hematopoietic maturation due to their abnormal genetic program. Objective: The objective of this study was to determine the aberrant phenotype in pediatric acute leukemia. Material and methods: This is an on-going cross-sectional study carried at the Department of Hematology and Transfusion Medicine, University of Child Health Sciences and Children's Hospital, Lahore, commencing on 1 October 2022 after IRB approval. The cases studied till December 2022 were 290 patients diagnosed with acute leukemias. Informed consent was obtained from patient's parents/guardians. Peripheral blood and bone marrow sample in EDTA vial were used for flowcytometric analysis by using FACSCanto II flowcytometer. The data was collected on a self-designed Proforma and analyzed by using IBM-SPSS V-23. Result: Among 290 cases of pediatric acute leukemia, the range of age was between 1 and 16 years with male to female ratio of 1.4:1. Out of total cases, 178(61.3%) had their imunophenotypic profile done with the help of Flowcytometric analysis on bone marrow aspirate while 112(38.6%) cases were analyzed using peripheral blood. Acute Lymphoblastic Leukemia (ALL) constituted 221(76.2%) cases and 69(23.7%) were Acute Myeloid Leukemia (AML). Out of the total, 32(11%) cases showed aberrant immunophenotype with the mean age was 7.9 ± 4.3 (Mean ± SD), the mean RBCs was 3.79 ± 1.2 × 10⁶/μL (Mean ± SD), the mean Hb was 10.1 ± 3.2 g/dL, the mean WBCs was 8.65 ± 6.4 × 10³/μL and the mean platelets count was 277 ± 281 × 10³/μL. The total aberrant antigens were 40(13.7%). The most common aberrant antigens were reported in B-ALL and the most common aberrant antigen was CD13 (62.5%). In AML the most common aberrant antigen was CD19 (55.6%) and in T-ALL the most common were CD117 and HLADR (26.6%). Conclusion: The most common aberrancy was seen in B-ALL cases followed by AML and then T-ALL. Aberrant immunophenotypic expression should be kept in mind while diagnosing the pediatric patients with acute leukemia as it may affect the prognosis of disease. Key words: Acute leukemia, ALL, AML, Imunophenotypic expression, aberrant markers, minimal residual disease.

¹The Children's Hospital, University of Child Health Sciences Lahore, Pakistan, ²FJMU Lahore, Pakistan, ³AIMC Lahore, Pakistan

Introduction: The Platelet Rich Plasma (PRP) is derived from venous blood and as the name indicates it has high quantity of platelets and can be used in enhancing the concentration of growth factors. Objective: The objective of the study was to compare the platelet count, platelet concentration/yield, residual Red blood cells (RBCs) and White blood cells (WBCs) counts in platelet-rich plasma (PRP) samples prepared from the single- and the double-centrifugation protocols. Materials and Methods: It was a Cross-Sectional study, conducted at the Department of Hematology & Transfusion Medicine, The Children's Hospital and UCHS, Lahore from October 2021 to January 2022 including 50 voluntary, healthy individuals of age 20–45 years of both genders, after taking informed consent. Complete blood count analysis of all participants was done initially by drawing 3 mL blood in EDTA vial. From all the participants, 20 mL venous blood sample was taken in syringes containing tri-sodium citrate and then shifted to harvest tubes. Group I comprised of PRP samples prepared by single- centrifugation method. While Group II samples were prepared by Double-centrifugation method consisting of soft and hard spin. The platelet, RBC and WBC counts in prepared PRP samples were determined by using automated SYSMEX XP-100 hematology analyzer. Platelet yield or Platelet concentration (%) was calculated for samples using formula. The data analysis was done using SPSS version 23. The full article is under publication process in PJMS. Results: The mean PRP platelet count in Group I was 594.6±157.4 × 10³/μL whereas in Group II was 923.06± 127.58 × 10³/μL. In Group I, the mean platelet concentration/yield in PRP was 175.75% ± 55.08% while in Group II, it was 276.78% ± 112.7%. Significant difference was observed between the platelet counts and platelet concentration/yields from the PRP samples of two Groups (p < 0.01). Significant difference between the WBCs count was also observed (p < 0.01) with higher WBCs in Group I PRP. Residual RBCs were almost same among two groups. Conclusions: The double centrifugation protocol resulted in higher platelet quantity and yield with less contamination by red and white blood cells than did the single centrifugation protocol for PRP preparation. So, double centrifugation method is beneficial in preparation of autologous as well as allogenic PRP.

Key words: Platelet-rich plasma; Platelet concentration; Centrifugation; Blood.

¹Indonesia Depok Indonesia, ²Indonesia Jakarta Indonesia

Introduction: Non-tuberculous mycobacterium (NTM) isan opportunistic pathogen for susceptible cases as in immunodeficiency state. Centers for Disease Control and Prevention (CDC) reported that NTM infection could be found in 10% of patients with Human Immunodeficiency Virus (HIV) presenting with symptoms resembling M. tuberculosis (MTB) to such an extent that NTM infection could be suspected in cases unresponsive to antituberculosis treatment. NTM infection may progress due to therapy inadequacy, and could disseminate mainly hematogenous to other organs such as central nervous system. CASE A 38-year-oldmale on admission withfever, suppurative cough, dyspnea, headache and altered consciousness. He had history of HIV infection on antiretroviral therapy(ART) since 2008, hepatitis B and C co-infections since 2017, and was diagnosed for acute myeloid leukemia (AML) in 2019. The patient subsequently diagnosed as AFB negative tuberculosis and being treated with anti tuberculosis regiment since April 2020. He fell in undernourished category with 14.5 kg/m² body mass index, physical examinations showed right orbital mass, oral candidiasis, rhonchi, and liver enlargement. Histology of bone marrow biopsy was concluded for AML. NTM was detected based on PCR TB CSF analysis. The patient had elevated blasts peripheral percentage >20%. His CD 4+ cell count was 357 cells/μL, and the results of Gram staining and sputum culture obtained isolates of Klebsiella pneumonia and Pseudomonas aeruginosa. There was an increase of procalcitonin level (7.84 ng/mL) and on the HCV RNA examination virus was detected (7.08 Log IU/mL. Conclusion: A 38-year-old male suffered from immunodeficiency condition due to HIV-AIDS causing opportunistic NTM infection, accompanied by AML which also exacerbates immunodeficiency. Patients also suffered from Hospital acquired pneumonia(HAP) and had hepatitis B and C chronic co-infection.

Key words: NTM, HIV, AML, HAP, Hepatitis

¹Laboratori Clinic Territorial ICS-IAS Girona, Spain, ²Group Neoma, Facultat De Medicina Girona, Spain, ³Institut Catal〠Oncologia Girona, Spain

Introduction: Toxoplasmosis is an infection caused by intracellular protozoan parasite Toxoplasma gondii (TG).It's typically asymptomatic in immunocompetent patients while in situations with compromised immune system is more likely to develop a severe primary infection including fever, confusion, headache, seizures, poor coordination within others. Transmission can be zoonotic, foodborne, congenital and in rare instances iatrogenic. Case report A 55-year-old male with hemoglobin SC disease presented with fluctuating fever, affectation of general state, dry cough and myalgias, two weeks after being discharged with a presumptive diagnosis of acute bronchitis. One month before, he went to Gambia and since then, the patient referred the symptoms started. Clinical examination revealed cataracts in eyes, bradyphrenia and reduced levels of consciousness with no signs of neck stiffness. Adenopathy's were not found. Abdominal computer tomography scan showed absence of spleen and magnetic resonance imaging scan was normal. Laboratory test results showed an hemolytic anemia together with a high percentage of erythroblasts, target cells, Howell-Jolly bodies and a non-negligible population of abnormal lymphocyte morphology (Hemoglobin 8.3 g/dL, Reticulocytes 132.7 × 10⁹/μL, Leucocytes 36.82 × 10³/μL, Erythroblasts 30%, IgG positive direct antiglobulin test, Lactate dehydrogenase 1764 U/L, Alanine aminotransferase 115 U/L, indirect bilirubin 1.11 mg/dL, haptoglobin 8.3 mg/dL, D-Dimer 10 248 ng/mL, Ferritin >2000 ng/mL, Reactive C Protein 0.92 mg/dL). Over the course of admission, his neurological status and eye disease continued to decline. Lumbar punction was performed with no alterations, bone marrow study was obtained with presence of some hemophagocytes and blood cultures were repeatedly negative. Malaria and trypanosomiasis were excluded with several thick blood smears as well as with molecular techniques. Differential diagnosis included Epstein-Barr virus, Cytomegalovirus, and HIV, HTLV D I/HTLV-II, VHS-1/VHS-2, Cryptococcus neoformans and TG. Serology testing for TG resulted positive for IgG and IgM with low avidity. Results were confirmed by PCR in blood and cerebrospinal fluid with positive and negative results respectively. The patient was finally diagnosed with Hemophagocytic syndrome secondary to acute toxoplasmosis infection with vitritis. Conclusions Blood smear may be not always a helpful diagnostic tool as in very reactive situations; lymphocyte morphology can lead to confusion between a hematological malignancy and reactive pathology. Acute toxoplasmosis can be manifested differently so it's important a broad differential diagnosis. In this case, clinical manifestation was mainly ocular and neurological. Appropriate testing with serological and molecular techniques is important for an appropriate diagnostic.

¹University of Witwatersrand Johannesburg South Africa, ²University of Amsterdam Amsterdam Netherlands

Introduction: Venous thrombo-embolism (VTE) is an important cause of maternal mortality and thromboprophylaxis is recommended in in pregnant and postpartum women with a history of VTE. The pharmacodynamics of low molecular weight heparin are altered during pregnancy as a result of the increased volume of distribution and enhanced renal clearance. Anti-Xa monitoring has been suggested to guide dose adjustments with advancing gestation. Nonetheless, there is limited efficacy and safety data with regular anti-Xa monitoring. Aim To assess anti-Xa adjusted enoxaparin thromboprophylaxis in pregnant women at high risk of VTE and its association with thrombosis and bleeding outcomes Methods A cohort study was conducted at the combined cardiac obstetric clinic in Johannesburg, South Africa. The clinical records of consecutive pregnant women, who received enoxaparin thromboprophylaxis for the prevention of VTE between March 1, 2017 and September 1, 2022, were reviewed. Pregnancies, with a personal history of VTE, were managed with anti-Xa adjusted enoxaparin antepartum from a median [interquartile range] of 12 [14] weeks’ gestation and for 6 [0] weeks' postpartum. Peak anti-Xa levels, at a mean ± standard deviation of 3.3 ± 0.4 hours, were monitored at regular intervals to achieve an anti-Xa level of 0.3–0.5 units/mL. Pregnancy-related VTE was objectively confirmed. Major, clinically relevant non-major and minor bleeding were defined according to the International Society on Thrombosis and Hemostasis Scientific Subcommittee. The study was approved by the Wits Human Research Ethics Committee, Medical (M-220207). Results In this cohort of 59 pregnancies, with a mean age of 33 ± 5 years, the live birth rate was 86.2%. Serial peak anti-Xa levels were available in 88.1% of pregnancies. With pregnancy progression, 19.2% required a dose adjustment in the second trimester and 25.0% in the third trimester to achieve the target range. VTE occurred only antepartum in 3.4%, 95% confidence interval (CI) 0.4-11.7 of pregnancies. Bleeding events occurred in 8.5%, 95% CI 2.8–18.7 of pregnancies of which 5.1% were classified as postpartum major bleeding. On univariate logistic regression analysis, no independent predictors of bleeding were identified. Conclusion The rates of thrombosis and bleeding with anti-Xa adjusted enoxaparin thromboprophylaxis were consistent with earlier reports in pregnant women at high risk of VTE. A randomised controlled trial is indicated to determine the efficacy and safety of anti-Xa adjusted enoxaparin as compared to fixed low dose enoxaparin thromboprophylaxis.

Department of Pathology, University of Iowa Hospitals and Clinics Iowa City, IA, USA

Introduction: The CS-5100 instrument utilizes a wavelength of 405 nm to determine fibrinogen activity based on optical clot detection. This method works well in icteric samples with normal or high fibrinogen. However, in icteric samples with concomitant hypofibrinogenemia, the instrument cannot detect clot formation through the darker-colored plasma, meaning the fibrinogen result is unreportable until resolution of the hyperbilirubinemia and/or hypofibrinogenemia. Given that our laboratory frequently encounters icteric samples (~25 patients per month with total bilirubin >15 mg/dL), we evaluated the performance of a higher wavelength (575 nm) for optical clot detection in the CS-5100 fibrinogen assay. Methods: The original fibrinogen assay parameters were reproduced to create a modified fibrinogen assay with the only change made being the wavelength from 405 to 570 nm. The performance of this new assay was validated by performing precision and comparison studies on two Siemens CS-5100 analyzers. In addition, we evaluated whether the new assay (570 nm) could reference the calibration curve from the original fibrinogen assay (405 nm). Results: Within-run and between-run precision were satisfactory (CV 3%–7%). Comparisons were run between the two fibrinogen assays using non-icteric (n = 18, total bilirubin <2 mg/dL) and icteric samples (n = 16, total bilirubin 12–32 mg/dL, direct bilirubin 10–30 mg/dL) spanning a fibrinogen range of 32–870 mg/dL. Measured fibrinogen agreed within 15 mg/dL or 10% in all comparison samples. Five samples were analyzed using both calibration curves and results agreed within 10%, demonstrating that a single calibration curve could be used for both original and modified fibrinogen assays. To trigger the alternate assay, laboratory staff are instructed via middleware rules to visually inspect samples with either a ‘Slight coagulation' or ‘No coagulation' error generated by the CS-5100 during fibrinogen measurement. For icteric samples, the alternate fibrinogen assay is manually ordered. Since implementation of this assay almost 2 years ago, it has been utilized approximately 10 times per month. Conclusions: An alternate wavelength of 575 nm can be used to precisely and accurately determine the fibrinogen activity in icteric samples on the Siemens CS-5100 instrument. This modified method is especially useful in patients with concomitant hyperbilirubinemia (total bilirubin >15 mg/dL) and hypofibrinogenemia (fibrinogen <100 mg/dL).

Aga Khan University Karachi, Pakistan

Establishment of Reference Intervals for Hemoglobin A, A2 and F in Pakistani Children using an Indirect Data Mining Approach Authors: Muhammad Shariq Shaikh (1), Sibtain Ahmed (1), Zeeshan Ansar (1), Shahnawaz Bayunus (1) Department of Pathology & Laboratory Medicine, Aga Khan University Hospital Pakistan. INTRODUCTION: Reference intervals (RIs) are important medical decision-making tools that help physicians in differentiating healthy from sick individuals. These intervals vary in between gender and age groups. Hemoglobinopathies are prevalent in Pakistan. Quantification of hemoglobin variants is pivotal in screening out hemoglobin disorders. Establishment of RIs using a direct approach is difficult, specifically in children. We chose a data mining method (indirect method) for the determination of Hemoglobin (Hb) A, A2 and F RIs. METHOD: Hemoglobin A, A2 and F measurements performed for both inpatients and outpatients aged birth to 18 years between January 2013 and December 2020, were retrieved from laboratory management system of clinical laboratories at the Aga Khan Hospital. The study population was approximately representative of the entire geographical distribution of the country. Hemoglobin subtype were measured on Bio-Rad VariantTM II Turbo. Reference intervals were computed using an algorithm developed and validated by the German Society of Clinical Chemistry and Laboratory Medicine's Working Group on Guide Limits. Assuming that non-pathologic samples follow a Gaussian distribution (after Box-Cox transformation of the data), an elaborate statistical process was used to isolate distribution of physiological samples from the mixed dataset and used to calculate reference intervals. For the validation of our results, we compared our derived RIs with those from the Hemoglobinopathy Diagnosis Text Book. RESULTS: A total of 75 419 specimens were analyzed during the study period. After the exclusion of patients with multiple specimens obtained during the study period, RIs were calculated for 63 332 children with stratification into age groups. A comparison of our 2.5th and 97.5th percentile results with those of established RIs from Text Book on Hemoglobinopathies demonstrated good agreement in between different age groups. CONCLUSIONS: This study supports data mining as an alternate indirect approach for establishing hemoglobin variants specifically in resource-limited settings. The results obtained are specific to studied population, instrument and reagent used. These intervals therefore, will assist in better clinical decision-making based on hemoglobin variant results.

Zhongda Hospital Southeast University Nanjing China

Background: The erythrocyte sedimentation rate (ESR) is a nonspecific inflammation indicator used to observe disease dynamics. The traditional ESR measuring method Westergren is time-consuming. There's a built-in ESR testing function in a new hematology analyzer Mindray BC-720, and we evaluated the clinical suitability of this method in measuring ESR from patient samples from different departments. Methods: We collected 100 blood samples each from the department of rheumatic immunology, orthopedics, and hematology in 2022 at the Zhongda Hospital Southeast University. ESR was measured and compared by BC-720 and Westergren methods. The clinical efficiency was also analyzed using the Westergren results as the reference. Besides, we selected samples with hematocrit (HCT) <35% to evaluate the performance of ESR from BC-720 in patients with possible anemia. We also studied the correlation between ESR measured by BC-720 with the course of the disease by measuring ESR and other indicators of 4 patients continuously. Results: The ESR value measured by BC-720 correlated very well with the Westergren method in the department of rheumatic immunology (r = 0.9736), orthopedics (r = 0.9681), and hematology (r = 0.9739), and no significant difference was found among them. The clinical efficienciesof the BC-720 ESR were 94% (rheumatic immunology), 94% (orthopedics), and 96% (hematology), using Westergren results as the reference. In the patients with HCT <35 %, the BC-720 ESR also showed good correlations in all 3 departments (r >0.96). The BC-720 ESR significantly increased with the severity of the diseases for the 4 patients in this study. Conclusion: The ESR function of the BC-720 auto hematology analyzer is a simple, quick, and reliable method to measure ESR in patients from rheumatic immunology, orthopedics, and hematology departments.

Yale University, USA

Advances in our understanding of the molecular pathogenesis of myeloid neoplasms have been a major theme in the updates to the classification of Acute Myeloid Leukemia, Myelodysplastic Syndromes, and Myeloprolferative Neoplasms. These updates have resulted in the International Concensus Classification (ICC) and the fifth edition of the World Health Organization Classification of Haemtolymphoid Tumours (WHO-HAEM5). This talk will focus on the recent updates to the classification systems with an emphasis on the molecular drivers of disease and how to integrate the molecular work-up into the diagnosis of myeloid neoplasms.

Emory University, Atlanta

Low VWF typically refers to persons who have a VWF activity or antigen between 30% and 50% associated with pathological bleeding. Recently an international effort convened to better define this group and improve access to our hemophilia treatment centers. Although diagnostically challenging, even moreso is how to manage acute bleeding events and in preparation for surgery. This short lecture will focus on changes to the VWD guidelines and characterizing those with low VWF or mild VWD.

Russian Research Institute of Hematology and Transfusiology Saint Petersburg Russia

Background: Polycythemia vera(PV) belongs to the group of Ph-negative myeloproliferative neoplasms and is characterized by thrombotic complications. It is known that a significant contribution to the development of prothrombogenic potential in patients with PV is made by the appearance of microparticles (MP) in the bloodstream, which appear as a result of intravascular activation of blood cells.MP expresses tissue factor (TF) and procoagulant phospholipids on their surface that can enhance prothrombogenic potential of blood plasma. Thus, unsafe and effective prevention of thromboembolic complications remains a major question in the management of these patients. Aim of this study to estimate the coagulation activity of microparticles in patients with polycythemia vera receiving the antiplatelet and combined (antiplatelet plus cytoreductive) therapy. Methods: The study included 44 PV patients receiving the antiplatelet (n = 22) and combined (antiplatelet plus cytoreductive) therapy (n = 22) and 30 healthy controls. Thrombin generation was measured in platelet-free plasma with Calibrated Automated Thrombinography (CAT). Following reagents were used: "FluCaâ€'kit", "PRP-reagent", containing rTF (1pM), "MP-reagent", containing negatively charged phospholipids (4 mcmol), with CTI (40 mcmol/ml). Endogenous thrombin potential (ETP, nM*min) was evaluated. STATISTICA 12.0 was used. The results are given as median (Me) and Q1-Q3. Mann-Whitney test was used to compare the studied groups. P<0.05 was considered statistically significant. Results: The data are shown in the table (Parameters of thrombinogramm in PV patients and healthy controls (Me; Q1-Q3)). PRP-reagent and MP-Reagent allow to estimate thrombin generation which depends on procoagulant phospholipids and TF, respectively. ETP (PRP-reagent) in patients receiving the antiplatelet therapy was significantly higher than in controls and tended to marked increase to the combined therapy patients (p < 0.076). ETP (MP-reagent)in all patientstended to increase compared to the control group (p < 0.4) but it wasn't any differences between both groups. Conclusion: Combined (antiplatelet plus cytoreductive) therapy in patients with polycythemia vera leads to decrease of thrombin generation caused by phospholipids of microparticles. Used therapy has no effect on hemostatic potential, which determined by TF of microparticles in patients with polycythemia vera.

Hospital Albert Einstein São Paulo, Brazil

Introduction: Lupus anticoagulant is one of the laboratory parameters used to diagnose antiphospholipid syndrome and is strongly associated with thrombotic events and gestational morbidities. Despite their importance, the standardization for this test is still complicated, leading to discrepant results depending on the method used. This study aims to analyze the different methods used to identify the presence or absent of the lupus anticoagulant. Methods: 81 citrated plasma samples, composed by 34 positive samples and 47 negative samples, double centrifuged and previously characterized according to the testing criteria of the most recent ISTH guideline, were separated. After being processed in the ACL TOP LAS 700 for lupus anticoagulant, they were stored in a −80°C freezer for later testing in the STA COMPACT. The tests were performed in each equipment using the algorithm, plasma pool, and reagents recommended by the respective manufacturer. The results were compared between them in their respective time in seconds and ratio in the APTT and DRRV screen assays and their results (positive or negative). Results: A moderate agreement with the reference methodology were obtained. The kappa value obtained for these tests were 0.488 and 0.508 when used protocol suggested by Werfen and Stago, respectively. However, when was compared both protocols, a slight agreement (kappa value: 0,189) was found, with more positive results in werfen protocols. The mean ratios of both the DRVV screen and APTT tests were close in both negative and positive reference tests with statistical significance between them (p-value of <0.001). In the intraclass correlation of means in DRVVT and APTT was demonstrated substantial correlation, though the confidence interval of the 81 samples was spaced out in the DRVV screen tests (0.284–0.924). In addition, the prolongation of the APTT test in the reference positive samples was more visible in the tests performed in the STA COMPACT (28 prolonged APTT). Conclusions: The tests showed a significant difference between the methods used compared to each other. Different sensibilities to lupus anticoagulant, commercial pool versus homemade and the order of tests (with or without mix step) could be the responsible for this discordance.

Russian Research Institute of Hematology and Transfusiology Saint-Petersburg, Russia

Background. COVID-19 demonstrates a high mortality because of rapidly progress to severe and critical cases with respiratory distress syndrome, coagulation dysfunction, multiple organ failure, etc. Therefore, early identification of the disease expansion is very important to the clinical diagnosis and treatment of COVID-19. The aim of our study was to investigate the value of some coagulation parameters–prothrombin test (PT), activated partial thromboplastin time (APTT), fibrinogen (Fg), D-dimer (DD), factor VIII activity (FVIII), ristocetin-cofactor von Willebrand's factor activity (vWF:RCo), von Willebrand's factor antigen (vWF:Ag), antithrombin (AT) and homocysteine (HCY)–in predicting of prognosis of COVID-19. Methods. These laboratory data were collected at hospitalization in 104patientswith COVID-19. Outcomes were divided into two types: hospital discharge and death. Statistical analysis was performed by non-parametric methods(median (Me), 95% confidence interval(95% CI) and Mann-Whitney U test, Statistica 12.0), p < 0.05 was considered statistically significant. Results. Coagulation profiles observed in both groups of patients reflect a hypercoagulability, but PT and AT were significantly lower, while DD and HCY levels – significantly higher in patients with poor prognosis, and no difference in the other parameters was observed (data shown in the table 1). * p < 0.005 with controls. Conclusion. The results of this study showed that hypercoagulation was present in patients with COVID-19 even at the early stage. Severe coagulation dysfunction is more likely to occur in critically ill patients. PT, AT, DD and HCY could serve as diagnostic indicators for disease outcome.

University of Utah and ARUP Laboratories

Von Willebrand disease (VWD) is a common bleeding disorder of platelet adhesion with six currently recognized subtypes. Laboratory diagnosis consists of an initial test panel including antigen, activity and factor VIII measurements, sometimes followed by further specialized testing. VWF activity/antigen testing ratios help to differentiate type 1 and type 2 disease, which is important for selection of proper therapy. Recommended ratio cutoffs differ by guideline, ranging from 0.5 to 0.7, with 0.7 commonly recommended. The ratio cutoff used affects the sensitivity and specificity for type 2 diagnosis. Variability in VWD due to underlying mutations and patient factors, as well as variability in VWF tests, impact the accuracy of ratios for VWD subtyping. This presentation will discuss use of activity/antigen ratios in the diagnosis and subtyping of VWD with a focus on technical aspects of the tests.

Special Coagulation Laboratory Rochester, MN, USA

Introduction: Lupus anticoagulants (LA) result in prolongation and inhibition of phospholipid based clotting time and shortens with additional phospholipid. The hexagonal phase phospholipid neutralization test (HPNT), a silica based assay, is one of four possible confirmatory tests recommended for LA testing. We evaluated a new, FDA approved kit for usability and performance against the predicate method on an optical based analyzer. Methods: Hex LA (CRYOcheck™ Hex LA™, Precision Biologic, Dartmouth, NS Canada) assays were performed on ACL TOP® Family 700/750 hemostasis instrumentation (Werfen, Bedford, MA). HPNT Delta (Delta) is calculated by comparing the clot times of patient plasma before (Start) and after addition of hexagonal phase phospholipid (Correct) at optical measurement of 671 nm. Precision studies were conducted on negative, weak, and positive lupus controls. Accuracy was assessed on waste samples submitted for LA panel testing to our laboratory (n = 52) and compared to results of Staclot LA (Staclot® LA 20, Diagnostica Stago, Asnières-sur-Seine, France)(predicate method). Results reported as: Hex LA delta(s): negative: <13.0, positive: ≥13.0; Staclot LA delta(s) negative: <9.0, positive: ≥9.0. Results: Precision data for three levels of control material across two lots of Hex LA demonstrated %CV ranging from 2.4 to 6.6 for Start/Correct, SD ranging from 1.2 to 5.6 for Delta. Method comparison observed 100% positive agreement (n = 23/52), 66% negative agreement (n = 19/52), 81% overall agreement (n = 42/52). Ten samples had postive Staclot LA and negative Hex LA; 8/10 were not suggestive of LA based on aggregate interpretation of the LA panel. 2/10 had borderline/positive IgM, and undectable IgG anticardiolipin antibody, undetectable IgM and IgG anti-beta 2 glycoprotein I antibodies, one cosistent with LA by two other paired test systems, the other not. Both had Hex LA deltas above manufacture normal range, but not elevated above laboratory verified interpretative cutoff. Conclusions: In our assessment, the Precision Biologic Hex LA kit was user friendly with preparation expediated by thawing instead of reconstituting of products. There are fewer kit components, with normal pool plasma added to the start and correct reagents. The kit demonstrated acceptable precision within and across lots. Accuracy assessments demonstrate good positive predictive value to the predicate method. Sensitivity and specificity was favorable, based on the panel of LA results. More observations and cautious interpretation are required in weakly positive LA.

Department of Laboratory Medicine, Peking University Third Hospital BeiJing, China

Background: Lupus anticoagulants (LA) increase the risk of thrombotic and obstetric events in patients with antiphospholipid syndrome than in those with other antiphospholipid antibodies. Anti-phosphatidylserine/prothrombin (aPS/PT) complex antibodies are thought to cause LA positivity. Therefore, we aimed to explore whether aPS/PT antibodies could prolong phospholipid(PL)-dependent clotting time and increase the risk of thrombosis or pregnancy complications based on LA positivity. Methods: We recruited 222 patients with positive LA and estimated their aPS/PT, anticardiolipin (aCL), anti-β2-glycoprotein I (aβ2GPI), and anti-β2GPI domain I (anti-D1) antibody (IgM and IgG) titersand PL-dependent clotting time [prothrombin time (PT), silica clotting time-screening (SCT-S), silica clotting time-confirmation (SCT-C), diluted Russell's viper venom time-confirmation (dRVVT-C), and diluted Russell's viper venom time-screening(dRVVT-S)]. Results: PT was longer in aPS/PT IgG-positive patients than in aPS/PT IgM-negative patients [12.1 (11.3, 14.4) s vs. 11.10 (10.6, 11.8) s; p < 0.001], while there was no significant difference between aPS/PT IgM-positive and IgM-negative patients [11.4 (10.9,12.4) s vs. 11.3 (10.7,11.9) s; P = 0.100]. Both SCT-S and dRVVT-S were prolonged in aPS/PT (IgG and IgM)-positive patients [IgG: 120.0 (75.0,158.0) s vs. 50.0 (40.0,70.0) s; IgM: 80.0 (57.8,125.8) s vs. 47.0 (38.0,67.3) s]compared to aPS/PT-negative patients [IgG: 48.0 (40.0, 62.0) s vs. 41.0 (36.0, 56.0) s; IgM: 46.0 (39.0, 52.0) s and 41.0 (36.0, 52.0) s]. The associations between aPS/PT IgG or IgM antibody titers and SCT-S or dRVVT-S were significant(P <0.001, P = 0.010, P <0.001, P = 0.002, respectively). SCT-C and dRVVT-C did not show any significant differences. The incidence of thrombosis in the aPS/PTIgG-positive group was much higher than that in the IgG-negative group (25.4% vs. 11.7%, χ² = 6.331, P=0.012). Likewise, the incidence of thrombosis was higher in the anti-D1- and aPS/PT IgG-positive patients than in the negative controls(40% vs. 14.3%, χ²= 3.934, P = 0.047).Furthermore, the aPS/PT IgG-positive group showed the strongest association with thrombosis [OR 2.584, 95% CI (1.213, 5.505)]. Conclusion::The aPS/PT antibodies prolonged PL-dependent clotting time, especially SCT-Sand dRVVT-S. In addition, the aPS/PT IgG-positive subtype increased the risk of thrombosis in LA positivity.

¹Department of Clinical Laboratory, Nanfang Hospital Guangzhou China, ²Department of Clinical Laboratory, West China Hospital of Sichuan University Chengdu China, ³Department of Clinical Laboratory, Tongji Hospital, Tongji Medical College of Hust Wuhan China, ⁴Department of Clinical Laboratory, The First Affiliated Hospital of Soochow University Suzhou China, ⁵Department of Clinical Research & Medical Affairs, Shenzhen Mindray Bio-Medical Electronic Co., Ltd. Shenzhen China

Background: Diagnosis of malignant hematological diseases often involves complicated tests such as morphology, immunology, cytogenetics, and molecular biology, which are complex and time-consuming. However, timely early screening for malignant hematological diseases is essential for effective treatments. In this study, an artificial neural network screening model for malignant blood diseases was constructed by blood routine tests and blood cell morphology parameters, and the effectiveness of the model was verified.

Methods: The venous blood samples of 1185 patients were collected from four tertiary hospitals nationwide as the training set. The patients were divided into non-malignant and malignant hematological diseases (lymphoma, acute lymphoblastic leukemia, acute myeloid leukemia, myelodysplastic syndrome, or myeloproliferative neoplasms patients) according to the clinical diagnostic information. Hematological and blood cell morphology parameters were collected using Mindray hematology and morphology analyzers, and relevant parameters were selected using the Filter method to construct an artificial neural network (ANN) model for screening malignant hematological diseases. Afterward, a testing set (n = 508) was utilized to evaluate the performance of the model. The accuracy, sensitivity, and specificity of this model to identify malignant hematological diseases were analyzed using the clinical diagnosis as the reference.

Results: Twenty-seven hematological and morphological parameters were selected for the ANN model construction. The ANN model achieved 90.96% screening accuracy for the training set, with sensitivity and specificity of 91.8% and 89.9%, respectively. For the testing set, the screening accuracy of the model was 89.51%, with a sensitivity of 85.33% and a specificity of 94.71%. Conclusion: The ANN model constructed in this study can effectively screen malignant blood diseases with peripheral blood samples, providing a novel and efficient auxiliary tool for malignant blood diseases diagnosis

¹Sysmex Corporation Kobe, Japan, ²Department of Clinical Laboratory, Tenri Health Care University Tenri, Japan, ³Department of Laboratory Medicine, Tenri Hospital Tenri, Japan, ⁴Department of Clinical Laboratory, Fukushima Medical University Fukushima, Japan, ⁵Medical Laboratory Division, Niigata University Medical and Dental Hospital Niigata, Japan, ⁶Department of Medical Oncology, Niigata University Graduate School of Medicine and Dental Sciences Niigata, Japan

Background Activated partial thromboplastin time (APTT) is used as a screening test for blood coagulation mechanisms in routine clinical laboratory tests, and the reasons as to why it may be prolonged vary. Some coagulation analyzers can output not only the coagulation time but also various parameters by Clot Waveform Analysis (CWA), which analyzes the reaction processes of APTT over time. These are used as additional information for inferring the cause of APTT prolongation. Sysmex's coagulation analyzers are capable of multi-wavelength CWA and are used for detecting abnormalities during APTT measurements, as well as for auxiliary measurements of difficult-to-measure samples. Objective We constructed a model to estimate the cause of APTT prolongation by processing multi-wavelength CWA waveform information with a convolutional neural network and evaluated its performance. Methods We included 669 samples with prolonged APTT, and the cause of APTT prolongation was confirmed based on medications administered and clinical information. The breakdown of samples was as follows: 100, Heparin; 258, Direct Oral Anticoagulants (DOACs); 105, Warfarin; 82, Lupus Anticoagulant (LA) positive; and 124, specimens with either decreased factor VIII, inhibitor positive, or decreased factor IX (FVIII/FVIIIi/FIX). CWA was measured using fully the Automated　Blood　Coagulation　Analyzer　CN-6000 and Revoheme APTT SLA. The performance of APTT extension cause classifications using the constructed deep learning system was evaluated by 10-fold cross-validation. Results The results (sensitivity, specificity, AUC) of prolongation of APTT using this model were as follows: Heparin (92.0%, 95.3%, 0.970), DOACs (86.8%, 95.1%, 0.959), Warfarin (99.0%, 95.0%,0.989), LA (61.0%, 95.1%, 0.934), and FVIII/FVIIIi/FIX (83.9%, 95.0%, 0.974). Conclusion The model constructed in this study was able to detect the probable cause of APTT prolongation with a sensitivity of 60% or more and a specificity of 95% using only the CWA waveform information obtained from APTT measurements.

¹JSLH Standardization Sub-committee for Blood Cell Counting Tokyo, Japan, ²Kansai University of Health Sciences Osaka, Japan, ³Nagasaki University Hospital Nagasaki, Japan, ⁴JSLH Standardization committee Tokyo, Japan, ⁵Nagasakiken Medical Association Nagasaki, Japan, ⁶Beckman Coulter Tokyo, Japan, ⁷BD Biosciences Tokyo, Japan, ⁸Keio University Hospital Tokyo, Japan, ⁹Shiga University of Medical Science Hospital Shiga, Japan, ¹⁰Kinki University Hospital Osaka, Japan, ¹¹University Hospital Kyoto Prefectural University of Medicine Kyoto, Japan, ¹²Osaka Medical College Hospital Osaka, Japan, ¹³Kyoto Prefectural University of Medicine Kyoto, Japan, ¹⁴Aichi Medical University Nagoya, Japan, ¹⁵Japanese Society for Laboratory Hematology Tokyo, Japan, ¹⁶The University of Tokyo Tokyo, Japan

Introduction: The internationally recognized flow cytometry reference method for reticulocyte percentage (RM-Retic) is established and used. It is important for practical use to confirm the convergence of the assay values for fresh blood calibrators measured by different laboratories. The Japanese Society of Laboratory Hematology (JSLH) executed assay value measurement in 12 laboratories to provide the target range for external quality assurance survey in Nagasaki prefecture. Methods: Fresh whole blood samples were collected from 2 healthy volunteers within blood bag containing CPDA solution, and EDTA-2K was added at a final concentration of 1.5 mg/mL. Aliquot volume (1.5 mL) of the samples were refrigerated and hand-delivered to each laboratory. The target assay ranges were determined by means of assay values measured by 13 flow cytometes (FACS CantoII 3, FACS Lyric 4, FACS Symphony 1 and Navios 5). The mean of 3 immunostaining tubes in 3 peripheral blood samples were used. The RM used the JSLH method which was validated by the CLSI H44 method. All examiners were trained using the RM procedure provided by JSLH. The samples were measured within 24 h after drawing. Target assay values were determined by means of assay values, and the target assay ranges as allowance interval (AI) was determined from %Bias. %Bias was calculated by Intra-individual biological variation (CVI) and between individual biological variation (CVG). The CVI and the CVG were used values in the biological variation database by the European federation of clinical chemistry and laboratory medicine (EFLM-BV). %Bias was calculated by 0.25 × (CVG² +CVI²)^1/2). The convergence of the target assay range as analytical variation (CVA) was compared with CVI. Results: The AIs of sample A for %reticulocyte were 1.52 (1.41–1.63). The AIs of sample B for them were 0.88 (0.82–0.95). The SD/CVA for %reticulocyte were 0.15/9.8% and 0.16/18.5%. All CVAs were over CVI (9.7%). The target assay ranges of sample A in each group of CantoII, Lyric, Symphony and Navios were 1.37–1.76, 1.65–1.65, 1.33–1.59 and 1.32–1.55. The ranges of sample B in them were 0.85–1.22, 1.11–1.11, 0.67–0.81 and 0.72–0.82. Conclusion: The convergence of the assay values for fresh blood calibrators measured by different laboratories were insufficient precision requirements of EFLM criteria of “desirable”. The AIs was determined by the mean of assay values and %Bias calculated CVI and CVG in the EFLM database.

Singapore General Hospital Singapore, Singapore

Introduction: Blasts with cuplike nuclear morphology have been reported in cases of acute myeloid leukaemia (AML) without maturation. The presence of cup-like blasts has been shown to be associated with certain distinct molecular and immunophenotypic features. The purpose of this study is to identify the frequency of such cases and elucidate the relationship between the presence of cup-like blasts and various immunophenotypic and molecular findings. Methods: We reviewed peripheral blood and bone marrow slides of patients diagnosed with AML without maturation at Singapore General Hospital over the past 10 years. Criteria for a positive case includes presence of at least 10% cup-like blasts, with cuplike nuclear invaginations covering at least 25% of the nuclear area. Clinical, immunophenotypic, cytogenetic and molecular findings were analysed and compared with a control group comprising cases of AML without maturation who do not have cup-like blasts. Results: We identified 15 AML patients with cup-like blasts and they accounted for 16% of all cases of AML without maturation. Cases of AML with cup-like blasts were significantly associated with a higher frequency of FLT3-ITD mutations (85% vs. 25%, p = 0.003), NPM1 mutations (92% vs. 25%, p = 0.0005), both mutations (85% vs. 13%, p = 0.0001), the lack of CD34 expression (92% vs. 25%, p = 0.0005) and lack of HLA-DR expression (100% vs. 63%, p = 0.03). There was a significant difference in the presenting white blood cell counts between the study and control groups (p = 0.0002). No significant relationship was found between the study and control groups with regards to cytogenetic findings. There were also no significant differences in age, gender and bone marrow blast percentage between the two groups. The positive predictive value for identifying AML with cup-like blasts for FLT3-ITD, NPM1 and both mutations are 85%, 92% and 85%. The negative predictive value for identifying AML with cuplike blasts for FLT3-ITD, NPM1 and both mutations are 75%, 75% and 88%. Conclusion: Our study corroborates existing data that cases of AML with cup-like blasts are highly associated with the presence of FLT3-ITD and NPM1 mutations, along with other immunophenotypic features. Identification of such distinctive cuplike nuclear morphology may be helpful in the prioritization of downstream laboratory workup of such cases of AML.

¹Diagnostic Laboratory Sciences and Technology Research Center, School of Paramedical Sciences, Shiraz University of Medical Sciences, Shiraz, Iran, ²Peyvand Pathobiology and Genetic Lab Complex, Shiraz, Iran, ³Research Center for Noncommunicable Diseases, Jahrom University of Medical Sciences, Jahrom, Iran

Introduction: The severity of the beta-thalassemia disease is mainly dependent on accumulation of the excess alpha-globin chains in red blood cell precursors which could impair erythroid cells. Mechanisms that reduce the accumulated alpha-globin chains could be beneficial to ameliorate the severity of symptoms in patients. Autophagy is an intracellular process for the removal of proteins and/or damaged organelles via lysosomes. mTOR pathway has been known as one of the key inhibitors of autophagy. Therefore, inhibiting the mTOR pathway could be a promising strategy to reduce the disease severity via autophagy-mediated clearance of alpha globin chains. Herein, we investigated the effect of metformin, an antidiabetic drug with mTOR inhibitory effects, on reduction of alpha-globin chains in nucleated red blood cells (NRBCs) isolated from beta-thalassemia major patients. Methods: NRBCs were isolated by Ficoll gradient centrifugation from blood samples collected from 17 beta-thalassemia major patients. NRBCs were cultured in RPMI 1640 complete medium and treated with different concentration of metformin for 4 days. Afterward, NRBCs were harvested and subjected to protein and RNA extraction. Changes in concentration of free alpha globin were measured by electrophoresis of proteins on Triton-Acid-Urea polyacrylamide gel. Furthermore, the expression level of autophagy-related genes (MAP1LC3B, ATG5, ATG7, and BECN1) were assessed by qRT-PCR. Results: Analysis of TAU gel electrophoresis results revealed that the concentration of free alpha-globin chains in NRBCs was significantly decreased by 9.5% ± 5.2%, 13.2% ± 8.6%, and 17.1% ± 10.5% after treatment with 0.4, 0.8, and 1 mM metformin. Besides, expression level of MAP1LC3B, ATG7 and BECN1 genes significantly increased in 1 mM metformin-treated NRBCs compared to untreated NRBCs (Fold changes: 8.2 ± 2.2, 77.1 ± 16.3, 137.5 ± 55.8, Respectively). The expression level of ATG5 gene also showed an increasing tendency which was not statistically significant (P value: 0.243). Conclusion: Findings suggest that metformin could be considered as a promising therapeutic agent for attenuating the disease severity in thalassemia patients due to its mTOR inhibitory-mediated autophagy activation effect.

Special Coagulation Laboratory, Mayo Clinic Rochester, MN, USA

Introduction: ADAMTS-13 cleaves ultra-large/high molecular weight von Willebrand factor (VWF) multimers (UL/HMWM) which accumulate in congenital or acquired deficiency of ADAMTS-13 and poses a risk for thrombotic thrombocytopenic purpura (TTP). The fluorescence resonance energy transfer (FRET)-VWF kinetic assay is critical for measurement of ADAMTS-13 activity in the diagnostic workup for TTP. We analyzed performance characteristics of a new ADAMTS-13 activity kit [ATS-13®ADAMTS13 Activity Assay 2.0 (Immucor, Waukesha, WI)] Ex492/Em518 compared to version 1.0 Ex345-350/Em440-450. Methods: ATS-13 2.0 was performed on the Perkin-Elmer Janus G3 liquid handling system and Agilent BioTek Synergy Neo2 microplate reader. Accuracy was performed on commercially available calibration materials and de-identified/archived plasma samples (n = 48) following IRB guidelines. Linearity was performed on four waste samples and the WHO first standard at a minimum of eight dilutions. Precision was established using three levels of pooled plasmas (assayed four times in duplicate, twice daily for five days). Limit of detection (LOD) and lower limit of quantification (LLOQ) was determined on kit diluent and low-level patient sample, each assayed 20 times per day for three days, on two different kit lots. Reference interval was based on 120 normal donor samples (60 male/60 female; ages 21–81 years). Analytical specificity assessed hemolysis, icterus, and lipemia (HIL) interference using HIL indices from the ACL TOP 750 preanalytical sample integrity check module (Werfen, Bedford, MA). Results: Quantitative accuracy showed an r² = 0.951, slope = 1.019, and mean % bias of −1.0% (Passing-Bablock regression). Linearity studies averaged an r² = 0.994 and slope = 0.989 with recovery within ±15% of assayed levels from 1.5% to 122.1% (122.1 IU/dL). Repeatability and within-laboratory CVs were 2.3% and 4.0% at ADAMTS-13 of 77.9%, 2.9% and 7.5% at ADAMTS-13 of 36.6% and 3.1% and 7.7% at ADAMTS-13 of 12.5%. LOD of 1.2% and LLOQ of 3.8% were established. The lower bound of the one-sided 95% parametric reference interval was 71.7%. Hemolysis did not exceed >25% loss of recovery up to H-index levels of 900 mg/dL. However, icterus and lipemia resulted in falsely decreased results at I-index and L-index levels >30 mg/dL and >4000 mAbs, respectively. Conclusions: Our data indicate that the performance characteristics of the ATS-13 2.0 assay are in accordance with International Council for Standardization in Hematology (ICSH) recommendations (PMID 32672897), and the assay can accurately and precisely measure ADAMTS-13 in the 5-10% activity range in our laboratory setting

¹Laboratory Oncology Unit, Dr. BRA-IRCH, AIIMS New Delhi, India, ²Department of Medical Oncology, Dr. BRA-IRCH, AIIMS New Delhi, India, ³Department of Biostatistics, AIIMS New Delhi, India

Introduction: Acute myeloid leukaemia (AML) is a malignant proliferation and abnormal maturation of myeloid progenitor cells. Wilms tumor 1 (WT1) gene is tumor suppressor gene and is a critical regulator of haematopoiesis, which encodes a zinc-finger transcription factor. The paired analysis of epigenetic mechanism of WT1 and its RNA expression has probably never studied in AML.In the present study, we aim to investigate the WT1 gene Realtime PCR based RNA expression & WT1 promoter methylation status and its impact in AML cases. Methods: After informed consent bone marrow (BM) and peripheral blood (PB) samples were collected from 112 AML cases (112 at diagnosis [day 0] and 105 after completion of induction chemotherapy [day 28] and 20 non-malignant samples were recruited as controls. WT1 gene expression and promoter methylation status were assessed during both intervals (day-0 & day-28) by performing real-time polymerase chain reaction (RT-PCR) and methylation-specific polymerase chain reaction (MS-PCR). Further, we explore the molecular and biological functions of WT1 by performing gene set enrichment analysis (GSEA) and the relationship of WT-1 expression with immune checkpoint analysis using the Sangerbox 3.0 database. Kaplan–Meier survival investigation was performed to estimate the prognostic significance of WT-1 gene in AML. Results: Of the 112 subjects studied, 73 were male (65.17%) and 39 were females (34.83%), out of these 97 (86.60%) cases showed overexpression of WT-1 gene at the time of diagnosis as compared with cases in complete remission (CR) remission or control samples (p = <0.001). Moreover, robust hypermethylation of WT1 promoter region was observed in 77 (68.75%) AML cases at the time of diagnosis as compared with patients in complete remission (CR) remission or control samples (p = <0.001). In all AML patients, WT1 expression and methylation levels were inversely correlated with normal haematopoiesis and positively associated with age, high marrow blast counts, M4 subtype, adverse risk cytogenetic, and inferior outcome compared to patients with low WT-1 methylation and expression levels. In gene set enrichment analysis (GSEA) we found WT-1 gene plays an important role in cancer produced as aberrations in molecular functions such as DNA-binding transcription factor activity, RNA binding, protein binding, and methylated DNA binding. WT1 also plays important role in cancer produced due to multiple molecular functions including negative regulation of cell growth, cell population proliferation, apoptotic process while positive regulation of DNA methylation. In immune checkpoint analysis WT-1 gene expression was positively correlated with CD28, CD40, CD44, CD48, CD80, CD70, CD27, CD86, and so forth. In Survival Analysis poorer overall survival was seen in patients having higher WT-1 gene expression as compared with the lower expression group. Conclusions: Overexpression and Hypermethylation of the WT-1 gene positively associates with the leukemic burden and poor prognosis. Thus, this gene can be considered a promising molecular marker for screening among high-risk cases (such as MDS) and MRD detection, and a target for developing novel therapeutic approaches against AML.

¹Sysmex UK Ltd Milton Keynes, United Kingdom, ²Sheffield Teaching Hospital NHS FT Sheffield, United Kingdom

Introduction: Haemolysis interference with coagulation tests is dependent on the reagents, detection systems, and through the biological activation of the haemostasis cascade. The use of paired matched haemolysed / non-haemolysed samples facilitates the true assessment of haemolysis interference. This is required to ensure validity and reliability of patient results. The aim of this study was to assess haemolysis interference on PT, APTT, Fibrinogen, D-Dimer and Thrombin Time (TT), using Siemens / Hyphen reagents on the Sysmex CN-6000 coagulation analyser. Methods: Blood samples were collected into 3.2% trisodium citrate and centrifuged at 2000 g for 10 min. Haemolysed samples were identified using the Sysmex CS-5100 haemolysis index. If a non-haemolysed repeat sample was received within 4 hours of collection of the haemolysed sample, residual plasma from both samples was aliquoted and frozen (−80°C). Paired matched samples were processed on the Sysmex XN-10 for haemoglobin quantitation (Hb), and on the CN-6000 for haemostasis assays. A reference range (RR) was established with 30 normal samples for PT, APTT, Fibrinogen, and TT. For D-Dimer, the 0.5 mg/L exclusion cut-off was utilised, and for PT INR, a therapeutic range of 2.0 – 3.0. Results: Shorter clotting times were observed in PT / APTT / TT when haemolysed, most pronounced in APTT, but also prolonged in a small subset. Variability between reagents for each assay was observed. Fibrinogen was either increased / decreased at the upper RR limit in haemolysed samples, but no difference was observed at the lower RR limit. Both D-Dimer reagents returned one false positive in haemolysed samples, and LIAPHEN^TM three false negatives, albeit two borderline (Tab. 1). No correlation between the level of Hb and observed differences were noted (r² <0.2). Conclusions: Haemolysis interference appeared variable, with decreased and increased clotting times / concentrations, irrespective of Hb. Interpretation with RRs demonstrated that haemolysis may influence patient diagnosis / treatment, most significantly for APTT, and less significantly for PT, Fibrinogen, D-Dimer, and TT. Automated haemolysis checks should be set to the lowest limit of identification and haemolysed samples for APTT should be rejected since the lowest levels of Hb can significantly impact measured values.

Clinical Laboratory, Core-Hematology Department. Germans Trias i Pujol University Hospital Badalona Spain

Introduction:. The new Mindray BC-780 is an automated hematology analyzer that combines blood cell analysis with erythrocyte sedimentation rate(ESR) detection. The aim of this study was to evaluate the analytical performance of ESR detection to verify its clinical utility. Methods:. The ESR performance was evaluated regarding repeatability and carry over (CR) of the Mindray BC-780 automated hematology analyzer, and compared this method with the modified Westergren method (VC-80: Ves-Matic Cube, Diesse), using 235 samples. Results:. Results obtained for imprecision and carry over met the technical requirements. The BC-780 ESR has good repeatability [standard deviation (SD) ≤1 mm/h, coefficient of variation (CV) ≤5%], and the CR was less than 1%. The ESR results measured by the BC-780 correlated well with the VC-80 analyzer (r = 0.952). The results of 235 samples for ESR detection are shown in Fig. 1. The correlation among the analyzers was good according to the results of the Spearman correlation coefficient.The regression analysis showed a proportional difference between these two different methods(y = −0.96 + 0.88x, r = 0.952) (Fig. 1-A). Fig. 1-B, and 1-C show the deviations of the BC-780 and the VC80. The Bland Altman plot for ESR of both analyzers showed that the BC-780 count tended to be lower as the average increased (Fig. 1-B). However, this bias did not lead to a clinically significant difference since the differences between the two methodologies did not translate into discrepant values ​​at the cut-off points used to make therapeutic decisions as it shows in the pair plot (Fig. 1-C) of ESR values measured by each platform, restricted to values between 0 and 150mm/h, where each pair represents one patient's sample. Conclusions. The Mindray BC-780 ESR showed high accuracy and good repeatability, which provided a faster, safer, and reliable technology to simultaneously measure ESR and hemogram parameters.

Clinical Laboratory, Core-Hematology Department. Germans Trias i Pujol University Hospital Badalona Spain

Introduction: Accurate measurement of PLT counts is essential, particularly in the management and follow-up of patients with thrombocytopenia. Despite the different technological advances, automated hematology counters that use only impedance method for platelet quantification present an important limitation to produce accurate and reproducible counts in patients with thrombocytopenia in the presence of interfering elements such as microcytic RBCs. Moreover, the presence of giant platelets, may produce falsely low PLT counts when only using impedance technology. The new BC-780 analyzer uses, for PLT counting, the sheath flow impedance method in conjunction with fluorescent technology and light scatter analysis, improving the accuracy of PLT count in the presence of these interferents, especially in low PLT counts. The aim of this study was to evaluate the analytical performance of this new hybrid PLT count method (PLT-H) to verify its clinical utility in situations such as macrothrombocytopenias. Method The PLT-H performance was evaluated regarding repeatability, linearity and carry over of the Mindray BC-780 automated hematology analyzer. For the evaluation of the platelet count interference study, we compared the PLT counts obtained by impedance (PLTI), optical (PLTO) and hybrid (PLTH) methods against the digital cell morphology system(DMS). ANOVA, Student's t test and Box-Plot were used to analyse the results from three different methods for platelet counting. Results: Results obtained for imprecision, linearity and carry over met the technical requirements. In the evaluation of interference by small RBCs, Student's t-test, the study correlation (Tab. 1-A) and the box plot analysis (Fig. 1-B) showed no significant differences and strong correlation between PLT-H count and DMS. The box plot (Fig. 1-B) representing differences in PLT-H counts regarding DMS showed only minimal scatter. The interference assessment by large platelets is shown in Tab. 1-B, which also stands out significant differences in mean PLT count between the impedance channel and DMS in thrombocytopenic samples with MPV>13fL. The best agreement was also obtained between PLT-H and DMS as shown in the box plot (Fig. 1-C) Conclusions The results of our study show how the PLT-H count in the presence of some of the most common interferents, provides values comparable to the optical count performed by our digital image analysis system, showing a robust correlation coefficient and minimal dispersion. This finding is important, because it demonstrates how impedance-based platelet counting methodology should be combined with complementary technology when low PLT counts in the presence of some interferents.

¹Department of Biomedical and Molecular Sciences, School of Medicine, Queen's University Kingston, ON, Canada, ²Clinical Pathology Department, Faculty of Medicine, Mansoura University Mansoura, Egypt, ³School of Baccalaureate Nursing, St Lawrence College, Kingston, ON, Canada, ⁴Department of Obstetrics and Gynecology, Queen's University Kingston, ON, Canada, ⁵Department of Oncology, Queen's University Kingston, ON, Canada

Background: Venous thromboembolism (VTE) is a common occurrence in cancer. Thrombosis risk varies according to several cancer and patient-related factors. Chemotherapy increases this risk up to 9-fold. Current VTE risk assessment models such as Khorana score and its modifications can identify patients at high risk when undergoing chemotherapy but have limitations. We aimed to assess the global coagulation profile and platelet changes following chemotherapy to determine whether hypercoagulable state can inform VTE risk assessment. We also aimed to examine the contribution of other cancer and patient factors to this risk. Methods: A single-center, prospective, longitudinal study. Patients planned to receive chemotherapy for their cancers were recruited. Blood samples were taken prior to and after chemotherapy cycles 1 and 2. Patients were followed for 6 months for any thrombotic events. Thromboelastography (TEG) parameters R, K, α, MA, LY30, and CI were evaluated, ADP-induced platelet aggregation was performed and platelet count and MPV were assessed at the same time points.Baseline demographics, cancer data, BMI, Khorana score and VTE risk factors were recorded. Paired t-test and ANOVA were used to assess the effect of chemo on those parameters. Results: A total of 33 patients aged 35–85 (18 breast, 11 endometrial, 5 ovarian cancer), with 79% postmenopausal and 21% metastatic diseases were evaluated to date. 53% of patients were under platinum-based chemotherapy. 22% of patients had comorbidities. Hypercoagulability was detected by TEG in samples 2, 3 based on a reduction in R, K, and an increase in α angle, MA, and CI, but these changes were not statistically significant over time. 43- 71% of patients had hypercoagulable TEG traces (Tab. 1). Khorana score (assessed in 27 patients) showed 6 were low risk, 19 were intermediate and 2 were high risk. The MA and CI significantly increased with intermediate and high-risk khorana scores when compared to the low-risk (p = 0.03 and p = 0.04 respectively).Despite being within normal ranges, there was significant reductions in platelet count (p = 0.001), size (p = 0.001) and aggregation (p = 0.001) after the first dose of chemotherapy. No thrombotic events were recorded. Conclusions: TEG can detect hypercoagulability after chemotherapy. TEG MA and CI can potentially help improve the application of Khorana score. Platelets show diverse changes after chemotherapy and unlike expected, aggregation is inhibited in this patient cohort. The effect of cancer chemotherapy on coagulation is complex and requires attention. Future work will focus on assessing these effects in each specific cancer type.

¹Pasteur University Hospital, Hematology Nice, France, ²Emile Müller Regional Hospital, Hematology Mulhouse, France

Background: Thrombotic thrombocytopenic purpura (TTP) is a rare thrombotic microangiopathy (TMA) caused by congenital or acquired severe deficiency of disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13 (ADAMTS-13), an enzyme cleaving von Willebrand factor (VWF). ADAMTS-13 deficiency leads to accumulation of ultra-large VWF multimers, and subsequent platelet aggregation and microthrombi formation. Aside clinical symptoms, TTP diagnosis is based on thrombocytopenia, hemolytic anemia with the presence of schistocytes, and severely reduced ADAMTS-13 activity (<10%). As TTP is a life-threatening condition, the triplet therapy, which include plasma exchange, rituximab, caplacizumab, plus corticosteroids must be given to patients with suspected TTP until results of ADAMTS-13 was obtained, and be pursued only in case of deficiency. Therefore, the rapid ADAMTS-13 testing is critical for early diagnosis and optimal management of patients with suspected acute TTP. Design of the study: To compare the cost-effectiveness of a strategy based on the on-site measurement of ADAMTS-13 activity using a rapid fully automated chemiluminescent assay (HemosIL AcuStar ADAMTS-13) versus its centralized measurement in a reference laboratory, we created two scenarios based on either the local measurement of ADAMTS-13 activity with a maximum time-to-result of 16 h (assay performed during daytime, 7/7) or its centralized measurement with a median time-to-result of 4 days (range: 2–8). The triplet therapy was hypothesized to be started on admission and stopped if ADAMTS-13 activity was >10%. In case of ADAMTS-13 activity <10%, an inhibitor screening would be performed at the reference laboratory, as this assay was not expected to be locally implemented. For the calculation, we considered the costs of reagent, shipping/handling of the samples to the reference laboratory, laboratory staff workload, reimbursement rates, and treatments administered in patients with suspected TTP until availability of the ADAMTS-13 result. Results: Based on our 2018–2021 activity, there was a mean of 60 prescriptions of ADAMTS-13 activity per year, most for the follow-up of TTP patients and for patients with non-TTP TMA. Seven were for the diagnosis of acute TTP, suggested by the French score, which was confirmed in four cases/year (mean values). On-site measurement of ADAMTS-13 activity was found to be cost-effective with a 16%-reduction of the costs compared to its centralized measurement in patients with suspected TTP during the first 4 days. Conclusions: Together with its short turnaround time (33 min) and full automation, our results suggest that the HemosIL AcuStar ADAMTS-13 assay could be an assay of choice to establish the diagnosis of acute TTP in emergency settings by rapidly measuring plasma ADAMTS-13 activity.

¹Hospital Universitario Galdakao Usansolo Galdakao, Spain, ²Hospital Universitario Araba Vitoria, Spain

Background: Myelodysplastic syndromes (MDS) are a constellation of different diseases sharing anemia in the great majority of cases, and this cytopenia defines these pathologies and their most dramatic clinical manifestations. We assessed the utility of complete blood count (CBC) and research parameters from the analyzer BC 6800 Plus (Mindray Diagnostics) to discriminate MDS-related cytopenia. Methods: Three hundred and ten samples were selected from the routine workload. 120 samples from healthy individuals were used established the reference intervals of research parameters, 60 patients diagnosed of MDS, 130 samples with cytopenia due to other diseases (aplastic anemia, other hematological malignancies). The CBC parameters and research use-only parameters in the cytopenic groups were compared using analysis of variance, considering P <0.05 significant. For the post hoc comparisons of outcomes between each pair of groups Scheffé correction was applied. Receiver Operating Curves analysis (ROC) was used to established the performance of individual and combined parameters for the detection of MDS. Results: In the MDS group Mean cell volume (MCV), reticulocyte hemoglobin (CHr), percentage of macrocytic and hypochromic red cells (Mac%, Hypo%), red cell distribution width (RDW) and hemoglobin distribution width(HDW) had increased values compared to the cytopenic and normal patients, while platelets, red and white cell counts (RBC, WBC) and NeuX (related to the cytoplasmic complexity of neutrophils) were lower (p < 0.001). NeuX, Mac%, HDW and RDW had higher Area under curves (AUC) 0.915 (95% CI 0.855–0.952), 0.798 (0.714–0.878), 0.777 (0.689–0.868), 0.765 (0.677–0.855) respectively, best Youden index at cut offs <318, >9.8%, >20%, >17.2% respectively. Combining these tests AUC 0.935 (95% CI 0.875–0.972), sensitivity 84.3%, specificity 89.8%. Conclusion: CBC and research parameters gain insights to the differential diagnosis of MDS among cytopenias of other etiologies.

¹Hospital Universitario Galdakao Usansolo Galdakao, Spain, ²Hospital Universitario Araba Vitoria, Spain

Background: V617F mutation of Janus kinase 2 (JAK2) gene is used in the diagnosis of Philadelphia-negative myelo-proliferative neoplasms (MPN) such as polycytemia vera (PV), primary myelofibrosis (PMF) and essential thrombocythemia (ET). Patients diagnosed with MPN often develop thrombotic events as an onset of symptoms or in evolution; the pathogenesis of thrombosis in MPN patients is complex but the presence of JAK2 mutation is an important risk factor. We aimed to investigate the prevalence of JAK2 V617F mutation in our MPN patients and its association with hematological and coagulation parameters. Methods: We retrospectively reviewed the records of 115 patients diagnosed with MPN, 27 PV, 28 PMF. 60 ET, based on the 2016 World Health Organization criteria, mean age 61 years (range 51-78). All the clinical and laboratory data were collected at the time of the diagnosis without any prior treatment, including complete blood counts (CBC), coagulation function and chromosome karyotype. CBC was analyzed with Sysmex XN counters. The parameters of coagulation included prothrombin time-International normalized ratio (PT-INR), activated partial thromboplastin time (APTT) and fibrinogen, were analyzed in Stago STA R Max3 analyzers. Allele specific real-time quantitative fluorescence PCR (AS-qPCR) was performed to detect JAK2V617F mutation. Student's t test and variance analysis (ANOVA) with post hoc Bonferroni test were applied to verify differences among groups; p values < 0.05 were considered statistically significant. Results: JAK2 V617F mutation was detected in 59.9% of the cohort, 74.0% of PV, 53.5% of PMF and 55.5% of ET cases. The entire cohort of MPN patients was divided into two groups according to mutation status. JAK2V617F mutation was associated with higher Hemoglobin (Hb), leukocyte (WBC) and platelet counts, p < 0.001. Hb mutant mean 182 (Standard deviation 11) wild type 152 (9) g/L WBC mutant 16.8* (1.3) wild type 11.6 (1.1) 10⁹/L Platelets mutant 498 (41) wild type 599 (68) 10⁹/L Significant difference of PT-INR between patients withJAK2V617F mutation and wild-type patients p < 0.001. At the time of the diagnosis PT-INR was higher in mutation positive patients compared to negative cases. PV 1.02 (0.48) versus 1.19 (0.11), p = 0.015 PMF 1.00 (0.85) versus 1.15 (0.11), p = 0.02 ET 1.05 (0.13) versus 1.15 (0.13), p = 0.025 The same trend was observed in Fibrinogen in PMF (290 ± 96 mg/dL vs. 419 ±69 mg/dL, p < 0.001) and ET patients (212 ± 55 mg/dL vs. 390 ± 88 mg/dL, p < 0.001). Conclusions: Thrombosis remain the most important complication in Ph–negative MPN patients that significantly affect prognosis and quality of life. Higher values in CBC parameters are associated to JAK2V617F mutation; abnormal coagulation parameters, such as PT-INR and fibrinogen, were also related to JAK2V617F mutation, it alters the coagulation function and might, at least in part, account for the increased vascular events risk associated with MPN. JAK2-V617F mutation is an important parameter in prognostic evaluation because increased risk of thrombotic complications must be considered in those patients.

¹Hospital Galdakao Usansolo Galdakao, Spain, ²Hospital Universitario Araba Vitoria, Spain

Background: The clinical course of COVID-19 range from mild cold-like symptoms to severe illness with acute respiratory distress syndrome and organ failure. The latter is often associated with a higher inflammatory response and platelet hyperreactivity; those patients present a high incidence of venous and arterial thromboembolic events. Immature or reticulated platelets (IPF) are the platelets newly formed by megakaryocytes and released from the bone marrow into the bloodstream. IPF refers to the proportion of immature platelets among the total number of platelets, are hyperreactive and show prothrombotic activity associated with arterial thrombotic events. Mindray BC 6800 Plus analyzer reports IPF as a part of complete blood count; in the reticulocyte channel platelets are enumerated by optical method which uses fluorescent reagent containing cyanine derived dye. IPF fraction is calculated, ranging 0.6%–5.6 % in healthy individuals. Our aim was to assess whether the level of thrombopoiesis, measured by IPF, is associated with the disease severity in COVID-19 patients. Methods: In the period from 1 December 2020 to 30 November 2021 acohort of 419 well-characterized COVID-19-patients was recruited. Venous blood samples were drawn from each patient at time points during hospitalization: on admission to the hospital, 3 days afterwards and on subsequent time point during hospitalization. Categorical variables were described as percentage and for continuous variables median and interquartile range (IQ) values were calculated. COVID-19 severity outcomes were defined as general ward (mild), intensive care unit (ICU) admission (severe), or in-hospital death. chi-squared analysis was used to test COVID-19 severity outcome associations with categorical variables including baseline disease conditions. Kruskal-Wallis test was applied in order to detect statistical differences among groups P <0.05 was deemed to be statistically significant Results: In our cohort median age was 64 years (25–75 IQ, 37–85), 64 % males; 342 patients stayed in general wards while 77 (18.4%) were admitted to ICU; 23 patients died during hospitalization (5.5%). The patients with a mild course showed normal platelet count 286 × 10⁹/L (IQ 168–388 × 10⁹/L) and a short-term increase in median IPF above the upper reference range value, peak value 7.7%. Severe ill patients had median platelet count 90 × 10⁹/L (IQ 44–141 × 10⁹/L), p < 0.001, and the highest peak value for IPF 10.1% (IQ 8.8%–12.6%) from the third day of hospitalization, p < 0.001. The median IPF was higher in ICU patients compared to patients with mild or moderate disease 6.8% (IQR 4.7%–10.6%) versus 3.6% (IQ 2.7%–5.6%) respectively, p = 0.001. Comparison of IPF between survivors and deceased patients revealed a clear difference with significant higher values in deceased (median 9.2%, IQ 4.5%–12.3%) than recovered (5.4%, IQ 2.8%–7.6%) patients, p = 0.001. Elevated IPF at presentation was predictive of ICU admission (P <0.001). Conclusions: in severe COVID-19 the number of platelets decreases while the newly formed immature platelets increases. This increase in IPF was more pronounced in critically ill patients. Elevated initial and peak values of IPF could serve as prognostic biomarkers for COVID-19 progression to severe conditions.

¹Hospital Universitario Galdakao Usansolo Galdakao, Spain, ²Hospital Universitario Araba Vitoria, Spain, ³H3L Consult Nuenen, Netherlands

Background Thalassemia trait and iron deficiency anemia (IDA) are the most common causes of microcytic anemia. These conditions are difficult to differentiate, since both genetic and acquired microcytic anemias display very similar clinical and laboratory features. The laboratory differential diagnosis of microcytic anemia starts with a complete blood count, supplemented with ferritin and other iron status parameters. For confirmation, thalassemia diagnosis requires hemoglobin analysis (HbA2 and abnormal Hb) and/or DNA mutation analysis. However, limited health-care resources are the rule rather than the exception in endemic thalassemia areas and these confirmatory assays may not be readily available. We apply artificial intelligence algorithm multilayer perceptron (MLP) for the evaluation of patients with microcytic anemia for discriminating between thalassemia trait and IDA. Methods We have retrospectively investigated a large, well-characterized patient population of 2664 patients with microcytic anemia: 1259 had iron deficiency (IDA), 1196 ‘pure' thalassemia trait (877 beta- and 314 alpha-thalassemia), 214 had thalassemia trait with concomitant iron deficiency, anemia of chronic disease or other diseases. Data sample was split randomly into a training set (n = 1853, 70%), and validation set (n = 811, 30%). In order to test the homogeneity of the laboratory findings in both cohorts the nonparametric Wilcoxon test was applied. Red cell count, Hemoglobin, Mean Cell Volume, Mean Cell Hemoglobin, Mean Cell Hemoglobin Concentration and Red Cell Distribution Width were used. Model performance was assessed using several measures of different types. The first type was represented by screening parameters according to a given threshold (sensitivity/recall, specificity, precision, accuracy and F1-score). The Youden index was used to select a good balance between sensitivity and specificity in the receiver operating curves. The F1 score was used in the precision-recall (PR) curves. The second type was the overall discrimination obtained using the area under curve (AUC); the value >0.8 was considered a good discriminatory indicator. Results For the training set Accuracy (Acc), 0.77; balanced accuracy (bAcc), 0.58; weighted F1 score (wF1), 0.75; AUC using one-vs-rest strategy of all classifiers using the derivation samples, 0.91. For the validation set Acc 0.81; bAcc 0.61; wF1 0.79; AUC 0.93 (95% Confidence Interval 0.920–0.945). Conclusions A correct diagnosis of thalassemia carriers is clinically relevant. MLP is a powerful machine learning method for accurate discrimination of thalassemia carriers from IDA patients using data from routinely available blood tests in clinical daily practice.

¹Hippokration General Hospital Thessaloniki Greece, ² P. Zafiropoulos SA Thessaloniki Greece

Introduction: Diabetes mellitus is a metabolic disorder, characterized by varying or persistent hyperglycemia, which induces several structural changes in erythrocytes, thus leading to alteration in their deformability and aggregation. Excessive availability of glucose within the cell yields an increase in HbA1c levels. The aim of this study was to investigate whether hyperglycemia corresponds to erythrocyte deformability and aggregation which may complicate the flow of these cells in microvessels. MATERIALS-Methods: Blood samples from 76 individuals were collected in tubes with EDTA as an anticoagulant. Erythrocyte deformability and erythrocyte aggregation were measured by RheoScan-D 300 system (RheoMeditech Inc., Korea), which is an ektacytometer using laser-diffraction technique with a microfluidic disposable channel. For Deformability, the range of the normal values varied between 0.285 and 0.360. On the other hand, for Aggregation, samples with values of <350 mPa were classified as normal. HbA1c was measured on HbNEXT HPLC analyzer (A.Menarini Diagnostics). The studied samples were divided in three groups: Group A (24 individuals) featuring normal glucose homeostasis (HbA1c<5.7%), Group B (24 individuals) with pre-diabetic patients (HbA1c:5.7–6.4%)] and Group C (28 individuals) with diabetic patients (HbA1c>= 6.5%). The Median, Minimum and Maximum values were calculated using the SPSS29 statistical software. In terms of statistically comparing among groups the Wilcoxon signed-rank test was employed, while Pearson's correlation coefficient was used for their correlation with HbA1c. The p-value < 0.05 was considered statistically significant. Results: The Median of RBC deformability was: group A:0.328 (0.3–0.56), group B:0.329 (0.24–0.57) and group C: 0.321 (0.322–0.37).The Median of RBC aggregation was: group A: 344.85 mPa (102.32–1062.8), group B: 369.25 mPa (192.5–793.3) and group C: 542 mPa (214.1–1977). A statistically significant increase of erythrocyte aggregation was observed in group C relative to group A (p = 0.005) and group B (p = 0.015) (Fig. 1). There was a decrease of erythrocyte deformability in group C which did not reach statistical significance relative to group A (p = 0.066) and to group B (p = 0.137). No statistically significant correlation between HbA1c and deformability /aggregation was detected for all groups. Conclusion: The measurement of erythrocyte aggregation may be used as a marker of severity of Diabetes mellitus and its complications due to the impairment flow of these cells in microvessels.

¹Norfolk and Norwich University Hospital Norwich, United Kingdom, ²Abbott Diagnostics Wiesbaden, Germany

Introduction: At the laboratories of Eastern Pathology Alliance, the hemocytometry is processed on 10 Alinity hq analyzers over three locations. Every Alinity hq analyzer has 2 dilution blocks resulting in a total of 20 channels that need to be controlled and aligned. For this, commercially available QC material, offered by the manufacturer, is measured twice daily. The error detection rate of this method is directly related to the frequency of QC analysis; this means that a large number of results could be affected by an analytical error before it is detected. Beside this, commercial QC material contains stabilized cells to increase the shelf-life, but these stabilized cells differ from fresh blood in terms of cell size, shape and deformability, and therefore may behave differently. The Alinity hq analyzers are equipped with an extensive moving average program. We investigated if moving average can be used as an additional tool to monitor the analytical quality of results. Methods: To observe the behavior of the moving average data at our three locations, each with a unique patient population, the moving average inclusion criteria for all Alinity hq analyzers were set according the following criteria: HGB 12.0–17.0 g/dL; RBC 3.80–5.50 × 10¹²/L; HCT 0–99%; MCV 70–110 fL; MCH 21.0–35.5 pg; MCHC 30.5–38.5 g/dL and RDW 0.00–15.0%CV. In total 120 batches with an average of 20 samples were collected over a period varying from 12 to 40 days. Moving average results were compared with the internal QC results (tri-level) over a period of 22 days. Results: Average and CV% for HGB was respectively 14.0 g/dL (13.9–14.1) and 0.9 CV% (0.8–1.3), for RBC 4.57 × 10¹²/L (4.50–4.62) and 1.0 CV% (0.8–1.5), for HCT 41.8% (40.6–42.5) and 1.1 (0.8–1.3), for RDW 12.8 CV% (12.5–13.0) and 0.8 CV% (0.5–1.8), for MCV 91.0 fL (89.8–92.2) and 0.8 CV% (0.6–1.0), for MCH 30.4 pg (29.9–30.9) and 0.7 CV% (0.6–1.1), and for MCHC 33.4 g/dL (32.2–34.4) and 0.5 CV% (0.4–1.0). CV% for the internal QC was in close relationship with the studied moving average parameters. Conclusion: We concluded that the moving average QC is a useful method for continuous monitoring of RBC parameters with added value to internal quality control.

¹Norfolk and Norwich University Hospital Norwich, United Kingdom, ²Abbott Diagnostics Wiesbaden, Germany

Introduction: At the laboratories of Eastern Pathology Alliance, the hemocytometry is processed on 10 Alinity hq analyzers over three locations. Every Alinity hq analyzer has 2 dilution blocks resulting in a total of 20 channels that need to be controlled and aligned. For this, commercially available QC material, offered by the manufacturer, is measured twice daily. The error detection rate of this method is directly related to the frequency of QC analysis; this means that a large number of results could be affected by an analytical error before it is detected. Beside this, commercial QC material contains stabilized cells to increase the shelf-life, but these stabilized cells differ from fresh blood in terms of cell size, shape and deformability, and therefore may behave differently. The Alinity hq analyzers are equipped with an extensive moving average program. We investigated if moving average can be used as an additional tool to monitor the analytical quality of results. Methods: To observe the behavior of the moving average data at our three locations, each with a unique patient population, the moving average inclusion criteria for all Alinity hq analyzers were set according the following criteria: WBC 4.00–12.2 × 10⁹/L; Neutrophil 37.0–78.0%; Lymphocyte 12.0–50.0%; Monocyte 0.00–14.5%; Eosinophil 0.00–8.60%; Basophil 0.00–1.50% and Immature Granulocyte 0.00–1.00%. In total 120 batches with an average of 20 samples were collected over a period varying from 12 to 40 days. Moving average results were compared with the internal QC results (tri-level) over a period of 22 days. Results: Average and CV% for WBC was respectively 7.3 × 10⁹/L (6.8–7.8) and 3.2 CV% (2.5–4.0), for Neutrophil 62.1% (60.4–64.3) and 1.9 CV% (1.4–2.4), for Lymphocyte 26.3% (24.2–27.7) and 4.4 CV% (3.2–5.3), for Monocyte 8.4% (8.3–8.5) and 2.7 CV% (2.0–3.5), for Eosinophil 2.2% (2.0–2.4) and 8.1 CV% (5.9–11.2), for Basophil 0.50% (0.39–0.56) and 9.0 CV% (6.5–12.3), and for Immature Granulocyte 0.06% (0.02–0.12) and 49.1 CV% (28.8–75.9). CV% for the internal QC was in close relationship with the studied moving average parameters. Conclusion: We concluded that the moving average QC is a useful method for continuous monitoring of WBC parameters with added value to internal quality control.

¹Norfolk and Norwich University Hospital Norwich, United Kingdom, ²Abbott Diagnostics Wiesbaden, Germany

Introduction: At the laboratories of Eastern Pathology Alliance, the hemocytometry is processed on 10 Alinity hq analyzers over three locations. Every Alinity hq analyzer has 2 dilution blocks resulting in a total of 20 channels that need to be controlled and aligned. For this, commercially available QC material, offered by the manufacturer, is measured twice daily. The error detection rate of this method is directly related to the frequency of QC analysis; this means that a large number of results could be affected by an analytical error before it is detected. Beside this, commercial QC material contains stabilized cells to increase the shelf-life, but these stabilized cells differ from fresh blood in terms of cell size, shape and deformability, and therefore may behave differently. The Alinity hq analyzers are equipped with an extensive moving average program. We investigated if moving average can be used as an additional tool to monitor the analytical quality of results. Methods: To observe the behavior of the moving average data at our three locations, each with a unique patient population, the moving average inclusion criteria for all Alinity hq analyzers were set according the following criteria: PLT 115–400 × 10⁹/L and MPV 6.0–11.0 fL. In total 120 batches with an average of 20 samples were collected over a period varying from 12 to 40 days. Moving average results were compared with the internal QC results (tri-level) over a period of 22 days. Results: Average and CV% for PLT was respectively 261 × 10⁹/L (252–270) and 2.8 CV% (2.2–3.4) and MPV 8.1 (7.8–8.3) and 1.5 CV% (1.2–2.0). CV% for the internal QC for the PLT was with 2.8% at a level of 73 × 10⁹/L; 1.9% at a level of 222 × 10⁹/L and 2.1% at a level of 485 × 10⁹/L in close relationship with the studied moving average PLT parameter. CV% for the MPV for the PLT of 73 × 10⁹/L was 1.3 CV%; 0.8 CV% at a level of 222 × 10⁹/L and 0.7 CV% at a level of 485 × 10⁹/L. Conclusion: We concluded that the moving average QC is a useful method for continuous monitoring of PLT parameters with added value to internal quality control.

All India Institute of Medical Sciences (AIIMS) New Delhi, India

Introduction: Inherited bone marrow failure syndromes (IBMFS) encompass a broad group of inherited disorders with cytopenias and a hypocellular bone marrow. Age at presentation varies from childhood to mid-adulthood. Absence of classical phenotypic features and/or a negative chromosomal breakage study (CBS) confounds accurate diagnosis; many such cases are mislabeled as immune aplastic anemia (AA). Genetic testing by next generation sequencing (NGS) helps in correct identification and appropriate clinical management of these patients. Methods: This is a retrospective study of 43 young (<30-years) patients presenting with features of bone marrow failure over a period of 2-years. Age group ranged from 4-months to 23-years. Clinical presentation included fatigue, bleeding, recurrent infections, growth abnormalities, and so forth. Physical abnormalities of IBMFS were seen in three patients. Peripheral blood evaluation showed persistent cytopenia of at-least one lineage. Bone marrow examination revealed a hypocellular bone marrow for age. Chromosomal breakage study was either negative or cell culture setup failed in these cases. NGS testing by a targeted gene panel (for genes associated with IBMFS) was studied in 38 cases and by clinical-exome sequencing in 5 cases. Results: Fifteen (34.8%) patients showed at-least one variant in the genes (Tab.-1) implicated in IBMFS (targeted panel was done in 13 cases and clinical-exome in 2 cases). Seven patients each showed one heterozygous variant; three patients showed homozygous variants; two patients showed a hemizygous variant and 3 patients showed heterozygosity for 2 different mutations. Among the seven patients with a heterozygous variant, autosomal dominant mode of inheritance was seen in three patients. Two patients showed a heterozygous variant with an autosomal recessive (AR) mode of inheritance. Another patient showed a heterozygous mutation in FANCC gene (AR mode of inheritance). RUNX1 mutation was seen in one patient which is associated with germline predisposition to myeloid neoplasm and pre-existing platelet disorder. Overall, NGS was helpful in twelve patients (12/43=27.9%) in correctly identifying an IBMFS. Variants in the genes involved in the Fanconi-anemia (FA) pathway were seen in five patients. Variants in 10 patients were seen in genes not related to FA. Conclusions: NGS is a useful method to identify the genetic abnormality of IBMFS in patients with non-classical features. Physical abnormalities may not be seen in all patients & CBS may be negative; many of these cases are falsely labeled as AA. Identification of a genetic basis helps to avoid anti-thymocyte globulin and screening family members as donors for stem cell transplant.

Jingzhou Hospital Affiliated to Yangtze University Jingzhou China

Background: Hemorrhagic fever with renal syndrome (HFRS) is a zoonosis endemic in eastern Asia caused by hantavirus. The reservoir host of HTNV is field mouse (Apodemus agraricus). The main manifestation of HFRS, including fever, bleeding and acute kidney injury. The disease progresses and deteriorates rapidly. Therefore, early and accurate differential diagnosis plays a key role in timely treatment and control of disease progression. Studies have shown that immature granulocytes and reactive lymphocytes in patients with HFRS are significantly increased in the advanced stage of the disease. In this study, the parameters and flag of the Mindray BC-6800 hematology analyzer of HFRS patients were analyzed, and a laboratory diagnostic model was established for rapid screening and differential diagnosis of HFRS. Methods: From January 2020 to June 2022, 88 patients with HFRS admitted to Jingzhou Hospital Affiliated to Yangtze University were enrolled as the case group. During the same period, 85 patients with acute kidney injury caused by bacterial infection were the control group. The clinical data of patients and the parameters and flag of BC-6800 hematology analyzer of two groups were collected, and the value of peripheral blood parameters for HFRS was analyzed by using the receiver operating characteristic (ROC) curve. Results: All the case group patients had fever and acute kidney injury. Compared with the control group, the lymphocyte count (Lym#) and lymphocyte% (Lym%) were higher (P < 0.0001) and lower PLT counts (P < 0.0001). ROC curve analysis shows that Lym# has the best diagnostic value (Lym#: 0.949, Neu%: 0.939, Lym%: 0.938, PLT: 0.813, WBC: 0.623 and Neu#: 0.542). Compared with the control group, Flags of “Blasts?” or/and “Abn. Lymph/blast?” or/and “Atypical Lymph?” and flag of “Immature granulocyte?” accounted for a relatively high percentage, reaching 96.6% (P < 0.0001) and 87.5% (P < 0.0001), respectively. Conclusion: The acute kidney injury patients with fever, showing abnormal hematology results including Lym# >1.65 × 10⁹/L, PLT <79 × 10⁹/L, presenting immature granulocyte flag and the flag of blasts, Abn. Lymph or Atypical Lymph, are suspected as high risk of HFRS.

¹Department of Haematology, Tan Tock Seng Hospital Singapore, Singapore, ²Department of Laboratory Medicine, Tan Tock Seng Hospital Singapore, Singapore, ³Lee Kong Chian School of Medicine, Nanyang Technological University Singapore, Singapore, ⁴Molecular Diagnostic Laboratory, Tan Tock Seng Hospital Singapore, Singapore, ⁵Yong Lin School of Medicine, National University of Singapore Singapore, Singapore

Introduction:: Paraprotein characteristics vary widely and is known to interfere with normal laboratory assays. We present an 82-year-old Malay male patient with a para protein which interfered with multiple routine laboratory assays due to its insolubility and precipitation as a gel in plain blood tubes and in blood tubes with common anti-coagulants. This necessitated the development of a novel method to solubilise the paraprotein in order to perform routine analysis like the CBC. Case report The patient presented with sudden left-sided weakness and reduced effort tolerance for 5 months. A MRI brain showed an acute infarct in the right frontal and parietal lobe. CBC was sent but the clotted sample clogged the automated cell counter. Left supraclavicular lymph node biopsy showed scanty aggregates of lymphoplasmacytoid cells consisting of a mixture of CD138+ plasma cells and CD20+/ PAX5+ B cells with monotypic expression of lambda immunoglobulin light chain. Bone marrow aspirate staining and flow cytometry could not be properly carried out because of the artefactual interference from the para protein. Serum electrophoresis was unsuccessful as the para protein could not go into solution. The trephine biopsy showed occasional para trabecular small lymphoid aggregates staining for CD3 and CD20 but negative for CD21 and Cyclin D1. Ki67 shows 5% proliferation index within the lymphoid aggregates with slight excess for lambda light chains. Cryoglobulin was negative. MYD88 mutation was also negative. Methods: To overcome the precipitation in anti-coagulated blood tubes, blood samples were immediately placed into a 37°C water bath and immediately transported to the lab. Glacial acetic acid (AA) and Owren Koller Buffer (OKB) were used to treat the sample (Fig. 1) to solubilise the para protein. Results: Using this method, an estimated manual CBC was performed. The hemoglobin was 9 g/dL, white blood cells ranged from 17.6–30.8 × 10⁹/L, and platelet count ranged from 132 to 312 × 10⁹/L. Serum protein and immune-fixation electrophoresis could be performed and showed an IgM lambda para protein. The serum free light chains (lambda: 319.9 mg/dL /kappa: 30.3 mg/DL) could also be performed. The patient refused plasmapheresis and was started on chemo-immunotherapy with bendamustine and rituximab. This led to reduction of the para protein and the resolution of para protein precipitation. There were also no further stroke complications. Conclusion: We present a case of an unusual para protein which required a novel solubiling method of using AA and OKB treatment, enabling the identification of the monoclonal band and the analysis of the patient's CBC.

¹Department of Hematology, Jiangsu Province Hospital and Nanjing Medical University First Affiliated Hospital Nanjing, China, ²Key Laboratory of Hematology of Nanjing Medical University Nanjing, China

Introduction: Real-time surveillance of the progression and evolution of chronic lymphocytic leukemia (CLL) has significant clinical implications as it could help inform patient management, yet realization lacks effective and convenient methods in clinical practice. Methods: This work standardized the analyzed area on blood films via an automated region of interest selection algorithm for the first time and built a convolutional neural network-based tool for feature extraction of lymphocytes in CLL patients. Results: Identifying lymphocytes in this work showed the highest recall (0.96) and F1 score (0.97). Based on this, we obtained multiple cell populations with different characteristics and morphology from CLL, which reflects different developmental stages of the disease to some extent. Correlation analysis substantiates these parameters from cell morphology have prognostic potential. The comprehensive end-to-end design makes it user-friendly and adaptable for a less-experienced population. Conclusions Our methods and analytical framework offer inspiration and a foundation for future work to characterize the effects of cell morphology on disease management. Future directions, including multi-center retrospective validation and prospective longitudinal studies, will help validate and establish the ultimate clinical utilities of this approach.

UT Southwestern Medical Center

This talk will focus on the latest classifications of myeloid neoplasms, genomic findings and the implications of these new classifications on treatment and patient outcomes. Myeloid neoplasms are a heterogeneous group of disorders that affect the blood and bone marrow, including acute myeloid leukemia and myelodysplastic syndromes. Recent advances in genetic technologies have enabled the identification of new subtypes of myeloid neoplasms, which are associated with different clinical presentations and outcomes. Implication of new classifications on diagnosis and treatment selection, as well as roles in identify patients that are at high risk for developing more aggressive forms of the disease and improve prognosis. The challenges and limitations of these new classifications and how we can work towards further improvements in the management of myeloid neoplasms will also be discussed.

Children's Hospital Los Angeles, USA

Childrens Hospital Los Angeles

The detection of residual acute leukemia following therapy is the single most prognostic factor identified for the monitoring of patients with these disorders. In North America, flow cytometry is the most commonly used technology for the detection of residual acute leukemia, although next generation sequencing is rapidly being incorporated into clinical practice. However, the detection of small abnormal populations of residual leukemic cells is more technically and interpretatively demanding than diagnostic evaluations. In this talk, the principles of detection of minimal/measurable residual disease by flow cytometry will be discussed.

University of Chicago, USA

FDA-approved viscoelastic testing (VET) methodologies have significantly changed in the last ten years, with the availability of cartridge-based VET. Some of these cartridge-based methodologies use harmonic resonance-based clot detection. While VET has always allowed for the evaluation of real-time clot formation, cartridge-based VET provides increased ease of use as well as greater portability and robustness of results in out-of-laboratory environments.

Here we review the use of VET in a variety of clinical contexts, including cardiac surgery, trauma, liver transplant, obstetrics, and hypercoagulable states such as COVID-19. As of now, high quality randomized trial evidence for new generation VET (TEG 6S, Hemosonics Quantra, ROTEM sigma) is limited. Nevertheless, the use of VET-guided transfusion algorithms appears to result in reduced blood usage without worsening of patient outcomes.

Future work comparing the new generation VET instruments and continuing to validate clinically important cut-offs will help move the field of point-of-care coagulation monitoring forward and increase the quality of transfusion management in bleeding patients.

¹Institute of Clinical Pathology and Medical Research, New South Wales Health Pathology, Westmead Hospital Westmead, Australia, ²Blacktown Hospital, Western Sydney Local Health District Blacktown, Australia, ³University of Sydney, Westmead Medical School Westmead, Australia

Introduction: KMT2A gene rearrangements are known to occur in both myeloid and lymphoid leukaemia. KMT2A amplification is rarer and generally reported to occur in only 1% of cases of acute myeloid leukaemia (AML) and is rarely reported in B-lymphoblastic leukaemia (B-ALL). To our knowledge, we present only the second case report of KMT2A amplification with corresponding TP53 deletion in adult de novo CD10 negative B-ALL. Methods: We present a 69 year old male who presented to the emergency department with symptoms of melena, coffee ground vomit, syncope and confusion, on a background of 2 weeks of abdominal pain. A full blood count showed he was pancytopenic with 17% circulating blasts. Flow cytometry and rapid fluorescence in situ hybridization (FISH) testing for BCR::ABL1 gene fusion and KMT2A rearrangement were performed on the peripheral blood to establish the blast cell lineage and further classify the leukaemia. Further FISH testing was also performed for TP53 gene deletion. Results: Flow cytometry testing showed that the blasts were positive for CD45(dim),CD19, CD34, CD38, CD15(variable) and HLA-DR. They were negative for CD3, CD4,CD5, CD7, CD10, CD13, CD20, CD33, CD56, CD64, CD117. Additional markers showed the blasts to be positive for intracellular CD79a, CD22 and to be negative for MPO and intracellular CD3, confirming that the diagnosis was B-ALL. The FISH showed KMT2A amplification with >20 signals per cell. There was no evidence of BCR::ABL1 gene fusion; however, there were 3~4 copies of BCR and 3 copies of ABL1 suggesting aneuploidy for chromosomes 9 and 22. Review of images captured favoured that the KMT2A amplification was localised to intrachromosomal regions and supported the possibility of clonal evolution. Conclusions: Further review of the literature found only 6 papers reporting a total of 7 cases of KMT2A amplification in B-ALL with only 1 adult B-ALL case without a history of previous chemotherapy. In the literature KMT2A amplification has been reported in association with a TP53 gene deletion in AML, but this has also been reported in ALL. FISH for TP53 confirmed that our patient also had a TP53 gene deletion and that a doubling clone exhibiting TP53 deletion was present. Multiplex PCR was negative for TCF3::PBX1, ETV6::RUNX1, KMT2A::AFF1 and BCR::ABL1 gene fusions. Unfortunately no further array or sequencing was performed due to the patient withdrawing from treatment and therefore the presence of other driver mutations for B-ALL were not assessed. This is a fascinating case of KMT2A amplification with associated TP53 gene deletion and clonal evolution in adult de novo precursor B-ALL.

¹Peking University First Hospital Beijing China, ²Aerospace Center Hospital Beijing China, ³Peking University International Hospital Beijing China, ⁴Peking University Shenzhen Hospital Shenzhen China, ⁵Beijing Miyun District Hospital, No. 383 Yang Guang Street Beijing China, ⁶Beijing Nuclear Industry Hospital Beijing China, ⁷Shenzhen Traditional Chinese Medicine Hospital Shenzhen China, ⁸Huazhong University of Science and Technology Union Shenzhen Hospital Shenzhen China

Background: Morphological identification of peripheral leukocytes is a complex and time-consuming task, having exceptionally high requirements for personnel expertise. This study is to investigate the role of artificial intelligence (AI) in assisting the manual leukocyte differentiation of peripheral blood. Methods: A total of 102 blood samples that triggered the review rules of hematologyanalyzers were enrolled. The peripheral blood smears were prepared and analyzed by Mindray MC-100i digital morphology analyzers. Two hundred leukocytes were located, and their cell images were collected. Two senior technologists labeled all cells to form standard answers. Afterward, the digital morphology analyzer unitized AI to pre-classify all cells. Ten junior and intermediate technologists were selected to review the cells with the AI pre-classification, yielding the AI-assisted classifications. Then the cell images were shuffled and re-classified without AI. The accuracy, sensitivity, and specificity of the leukocyte differentiation with or without AI assistance were analyzed and compared. The time required for classification by each person was recorded. Results: For junior technologists, the accuracy of normal and abnormal leukocyte differentiation increased by 4.79% and 15.16% with the assistance of AI. And for intermediate technologists, the accuracy increased by 7.40% and 14.54% for normal and abnormal leukocyte differentiation, respectively. The sensitivity and specificity also significantly increased with the help of AI. In addition, the average time for each individual to classify each blood smear was shortened by 215 s with AI. Conclusion: AI can assist laboratory technologists in the morphological differentiation of leukocytes. In particular, it can improve the sensitivity of abnormal leukocyte differentiation and lower the risk of missing detection of abnormal WBCs.

¹Department of Pathology, King George's Medical University Lucknow, ²Department of Internal Medicine, King George's Medical University Lucknow, ³Department of Gastroenterology, Sanjay Gandhi Postgraduate Institute of Medical Sciences Lucknow, ⁴Department of Neurology, King George's Medical University Lucknow, ⁵Department of Paediatric Surgery, King George's Medical University Lucknow

Introduction: Vaccination against COVID-19 has reduced the severity of the disease, yet we are still facing waves of COVID-19 infection like the recent surge in China. Demand for booster dose is brought up with each wave of COVID, especially in vulnerable population like elderly and immunocompromised population. Studies have monitored the antibody levels at different time interval to check for immune response in healthy individuals. Equally, T cell response has also been studied in healthy individuals. But there is paucity of data in immunocompromised individuals for sustained immune response specially with regards to T cell-based immunity. We studied different subset of T cell memory cells along with antibody titer at baseline and at 1 year duration in vaccinated healthy individuals and patients with liver cirrhosis at one year to observe any difference in immune response. Material and methods This was a prospective observational study including 36 adult heathy heath care workers (HCW) and 19 patients with underlying liver cirrhosis. 10 non-vaccinated healthy individuals were also included in the study. Memory T Cell subset analysis along with Anti-SARS-CoV-2S antibody & neutralizing antibody was assessed at baseline and one year in all three group by flow cytometry. Results: The patients with liver cirrhosis had significantly depleted T (CD3+T) cells as compared to healthy HCW and unvaccinated individuals though there was no difference between CD4+T and CD8+T cells. Significant difference was noted in various memory subsets on CD4+T and CD8+T [effector memory RA (P = 0.141, P = 0.000), effector memory (P = 0.000, P = 0.00), central memory (P = 0.000, P = 0.00), stem cell memory (P = 0.009, P = 0.08) and naïve (P = 0.000, P = 0.02)] in all three groups respectively. However, on post-hoc analysis both vaccinated HCW and vaccinated cirrhotic individual showed no difference in CD4 and CD8 memory T cell subsets. Figure flow plot showing different subset of memory cells on CD+ and CD8+T cells. Conclusion: Both heathy individual and cirrhosis patients have adequate number of CD4+ and CD8+ T memory cells irrespective of antibody levels and there is no significant difference between memory T cell subset in cirrhosis patients; hence they do not require additional doses of vaccine as compared to healthy individuals.

¹Vancouver General Hospital Vancouver, BC, Canada, ²University of British Columbia Vancouver, BC, Canada

Introduction: Treatment with clozapine requires regular monitoring of the white blood cell and absolute neutrophil counts for optimal patient care. Clients often live in marginalized situations resulting in low compliance with monitoring. To alleviate the strain on the clients, there is a need for a point-of-care (POC) hematology device that provides rapid and reliable results both in general and in cytopenic samples and is suitable for outpatient setting. We evaluated the CSAN PRONTO^TM device for WBC and ANC parameters at the ranges 1.1 to 21.3 × 10⁹/L and 0.4 to 15.8 × 10⁹/L respectively. Methods Three point-of-care devices CSAN PRONTO^TM were compared with the Sysmex XN9000 in 46 routine venous samples and 3 levels of commercial quality controls, focusing on white blood cells (WBC) and absolute neutrophil count (ANC). Results The CSAN PRONTO^TM was less precise compared to the acceptance criteria set for hematology instrument XN9000, with a CV% 3.0–3.7 for WBCs and 3.6–8.4 for ANCs. Correlation coefficient for WBC was r²>0.99 and ANC r²>0.98 for the entire range. Between instrument correlation was >0.99. Between run and within run precision of the instruments was well within manufacturer targets for both WBC and ANC (CV% <7.5% and <20% respectively). Linearity r² for both parameter was >0.99 in range 1.0–25.0 × 10⁹/L except for positive ANC bias in monocytosis and eosinophilia. Sysmex suspected flags, marked thrombocytosis, high/low hemoglobin, icteric, lipemic samples showed no interference. Conclusions The CSAN PRONTO^TM provides rapid and accurate WBC and ANC results, to aid clozapine treatment. The device is easy to operate and can measure both venous and capillary samples. Therefore, the CSAN PRONTO^TM is well suited as POC for nurse practitioners and physicians in an outpatient setting.

Nihon Institute of Medical Science Saitama, Japan

Introduction: Sepsis is caused by pathogen infection, where neutrophil extracellular traps (NETs) are formed to prevent the systemic spread or tissue invasion of invading microorganisms. Currently, presepsin, procalcitonin, interleukin-6, CRP, and so forth, are employed for the early diagnosis of sepsis. We focused on the component concentration value in a platelet absorption cytogram with ADVIA 2120i hematology analyzer, we investigated whether the extracellular release of neutrophil granules associated with NETs formation could be reflected as changes in component concentrations. Methods: Polymorphoprep (Axis Shield), a blood cell separation solution, was used to isolate polynuclear cells. NETs, induced by PMA (phorbol 12-myristate 13-acetate) reaction at 37 for 1–4 h, were measured complete blood count (CBC) with ADVIA. Using the Retic channel data stored in ADVIA, platelet absorption cytogram were analyzed with the FCS support system (SAICE Corporation). On the Y-axis, which shows the component concentration value, Intact areas were set at 60–100, pre-NET areas, indicative of the preparatory phases before NET formation, at 30–59, and NET areas at 0–29. The ratio of cell number in each area to the total cell number in three areas was determined by analysis with the FCS support system. Results: Before NET induction, polynuclear cells were detected in areas with high component concentrations. However, after NET induction, they gradually shifted to those with low component concentrations. This change was suppressed by the addition of an inhibitor for NET induction. Subsequently, three areas, that is, Intact, pre-NETs, and NETs, were statistically analyzed, demonstrating a significant decrease in Intact (%) and significant increases in pre-NETs (%) and NETs (%) 2 h after induction, compared to the untreated (UT) control. In addition, increases in Intact (%) and pre-NETs (%) and a decrease in NETs (%) were significant after the addition of an inhibitor for NET induction. Conclusion: Analysis of component concentrations in the platelet absorption cytogram of ADVIA suggested that NETs formation could be predicted only by CBC measurement, which may be useful for early detection of sepsis.

¹R&D, Shenzhen Mindray Bio-Medical Electronics Co. Shenzhen China, ²Xiangya School of Medicine, Central South University Changsha China

Introduction: Accurate platelet counting is crucial for disease diagnosis and treatment. The impedance PLT counting (PLT-I) approach could be interfered with small non-PLT particles in the blood. The fluorescent PLT count methods provide a more accurate but rather costly option. Thus, it is highly desirable to develop an anti-interference PLT counting method without increasing the cost. In this study, we introduced a hybrid PLT counting (PLT-H) strategy that combines the impedance and the conventional WBC differential (DIFF) channel in the Mindray BC-700 series hematology analyzer. We explored the principle behind PLT-H and evaluated the accuracy of this new technology. Methods: In our preliminary studies, we found that the common interferences affecting the PLT-I count were larger than 10 fL, which was then utilized as the borderline PLT size to measure the PLTs in the impedance channel. We selected blood samples with known interferences and measured the PLT count with both impedance and CD41/CD61 immunoplatelet (immunoPLT) reference methods, and then the PLT counts were compared by size to investigate the accuracy of PLT-I. We also explored the possibility to count PLTs larger than 10 fL in the DIFF channel. The morphology of PLTs treated with the hemolytic agents was observed with phase contrast and electron microscopies. And then the PLTs were measured and compared by size with both DIFF channel and immunoPLT. Finally, the accuracy of PLT-H was assessed by comparing it with the reference method in blood samples with interfering factors. Results: We found that there was no significant difference between PLT-I and immunoPLT for PLTs smaller than 10 fL in the interfering blood samples. After treatment with the hemolytic agent, PLTs preserved intact cellular structures, while RBCs were lysed into tiny particles smaller than 7.5 fL. The DIFF channel scatter plot showed a clearly defined region of PLTs, indicating PLTs could be counted in the DIFF channel. For PLTs larger than >10 fL, the PLTs counts from the DIFF channel showed a good correlation with immunoPLT. Furthermore, total PLTs measured with PLT-H in the BC-700 series hematology analyzer correlated well with the reference method (r = 0.9911) for blood samples with varying interfering factors. Conclusions: PLT-H provides a novel and accurate strategy by combining traditional PLT counting methods. It is not only an anti-interference PLT counting method but also economic-friendly since it did not add additional channels and reagents, which offers another option for hematology laboratories.

¹Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg Erlangen, Germany, ²Siemens Healthcare GmbH, Center for Innovation in Diagnostics Erlangen, Germany, ³Siemens AG Munich, Germany, ⁴Hospital Clinic de Barcelona Barcelona, Spain

Introduction: Hematological diseases, including leukemia, lymphoma, etc., are associated with White Blood Cell (WBC) disorders. Therefore, precise classification of WBCs is critical in peripheral blood (PB) smear analysis. However, the traditional staining process on the smears is time-consuming and expensive. Additionally, the chemical residuals are detrimental to the operators and the environment. In this work, we develop a two-stage deep learning (DL) pipeline for the classification of 14 different types of unstained WBC. Methods: PB smears of 233 different patients were collected from two different hospitals. These PB smears were microscopically digitalized before and after staining and a dataset consisting of both normal and abnormal WBC digital images was acquired. The stained cell images were initially annotated by two clinical experts. These annotations were then transferred to the unstained ones by alignment. A 14 different classes of blood cells have been annotated. The 14 types of WBC images are as follows: Basophil granulocyte (BA), blast (BL), eosinophil granulocyte (EO), immature granulocyte (IG), neutrophil granulocyte (NE), lymphocyte (LY), abnormal lymphoid cell (ALC), monocyte (MO), reactive lymphocyte (RL), plasma cell (PC), nucleated red blood cell (NRBC), giant platelet (GPLT), smudge (SMU) and artifact (ART). The total number of unstained images is 339 885. A 70%, 17%, and 13% of the total dataset were used for training, validation, and testing, respectively. As shown in Fig. 1, a patient-wise classification strategy with a transformer was designed to improve the classification rate using a pre-trained ResNet-18 as the feature extractor. This was performed to potentially mimic how pathologists classify by comparing different cells of the same patient to classify the normal and pathological cells. Results: Performing the evaluation on the testing set for both ResNet-18 and transformer models, mean F1 weighted scores of 90.33% and 91.77% are achieved, respectively. Thus, transformer approach with patient-wise method outperforms ResNet-18 by 1.44%. Fig. 2 shows the density confusion matrix of the best model with patient-wise transformer strategy. Even though the classification of normal WBCs is robust, the identification of some abnormal types still requires improvement. Conclusion: The proposed two-stage DL method shows promising results on unstained WBC classification. The transformer approach improves the classification performance on top of conventional ResNet-18 using patient-wise strategy.

University of Minnesota

External quality assessment (EQA), also known as proficiency testing, provides laboratories insight into the accuracy of their testing. Providing EQA material for platelet function testing creates unique challenges due to the need for fresh platelets. EQA providers have used multiple strategies to assess phases of platelet function testing. The North American Specialized Coagulation Laboratory Association (NASCOLA) in conjunction with ECAT Foundation offers post analytical surveys for platelet aggregation and dense granule testing. This session will discuss the knowledge learned from these platelet surveys.

¹The First Affiliated Hospital, Sun Yat-sen University Guangzhou China, ²The First People's Hospital of Foshan Foshan China, ³Zhongda Hospital Southeast University Nanjing China, ⁴Department of Clinical Research & Medical Affairs, Shenzhen Mindray Bio-Medical Electronic Co., Ltd. Shenzhen China

Background. Every hematology testing laboratory should establish appropriate hematology criteria and minimize the review rate without compromising the quality of the blood tests. The International Consensus Group for Hematology Review developed guidelines for action following automated hematology analysis in 2005. However, widespread implementation has been challenging due to its high review rate. This study aims to use artificial intelligence algorithms to develop a neural network hematology review model based on scatter plot features for accurate screening of samples requiring manual review. Methods:. We collected 8743 test samples from routine laboratory work for AI algorithm development and optimization and another 5809 samples for algorithm validation (AI group). These samples were also estimated using the International Consensus Group for Hematology Review guidelines (consensus group). All samples were run on the BC-6800 Plus hematology analyzer and validated by blood smears on the MC-80 hematology cell morphology analyzer. The blood samples were classified as “positive” or “negative” based on the ISLH criteria for positive smears. The slide review rate, false negative (FN) rate, and false positive (FP) rate were analyzed. Results:. Of the 5809 samples collected for validation, 903 were positive (15.5%). The AI group showed a significantly lower FP rate (7.8%) and a significantly lower review rate (23.3%) than the consensus group (19.9% and 35.4%, respectively, P <0.01). The AI algorithm was developed based on the features extracted from the scatter plot, and it only recognizes blood samples with more significant abnormalities, leading to fewer FP samples. The FN rate was slightly higher for the AI group (3.1% vs. 1.2%, P <0.01). We did a detailed analysis of the FN samples, and no samples with blasts or promyelocytes were missed in either group. The FN samples missed by both groups include samples with low myelocytes and metamyelocytes (<5%) (consensus: 59 samples vs. AI: 151 samples) and low NRBCs (<3%) (consensus: 4 samples vs. AI: 19 samples). A 7 samples with 5-10% of myelocytes and metamyelocytes were missed by the AI but not the consensus group. None of the FN samples had any impact on medical decision-making. Conclusion:. The scatter plot feature-based neural network review model is more advantageous for screening abnormal samples. It did not miss any clinically critical abnormal cells (e.g., blasts), which significantly improves manual review efficiency while ensuring test quality. It is a valuable hematology review model for hematology laboratories with large sample sizes.

¹Department of Laboratory Medicine, The First Affiliated Hospital, Sun Yat-sen University Guangzhou China, ²Department of Clinical Research & Medical Affairs, Shenzhen Mindray Bio-Medical Electronic Co.Ltd. Shenzhen China, ³Yunkang School of Medicine and Health, Nanfang College Guangzhou China

Background: Sepsis is an organ dysfunction syndrome caused by a dysregulated host response to infection. Due to the lack of specific diagnostic indicators, sepsis seriously affects the early treatment and causes high mortality. The new-generation hematology analyzers could obtain the characteristic information of the peripheral blood cell volume, cell internal complexity, and nucleic acid content, and then automatically outputs the quantitative detection parameters of cell morphology, thereby reflecting the infection degree of the body. In this study, a parameter was developed based on the morphological parameters provided by the hematology analyzer to screen sepsis. Methods: The training set included 515 patients admitted to the Emergency Department of the First Affiliated Hospital of Sun Yat-sen University that met the study's eligibility criteria. The patients were divided into sepsis (136 patients) and non-sepsis group (379 patients) based on the sepsis 3.0 criteria and the patient condition on admission day. The morphological parameters of WBCs were collected and analyzed with the binary logistic regression analysis. A diagnostic hematological parameter of sepsis (HPS) was developed based on the training set and evaluated with a validation cohort of 231 patients. The diagnostic performance of HPS was analyzed with ROC curves and compared with C-reactive protein (CPR) and procalcitonin (PCT) in sepsis diagnosis. Results: In the training set, the age of the sepsis group patients was 64 years (61–67), and that of the non-sepsis group was 61 years (58–63), with no significant statistical difference (P > 0.05), indicating that the difference was balanced and comparable. Binary logistic regression analysis showed that Neu_Y (the fluorescent light of neutrophils) and Mon_XW (particle distribution width of monocytes in the side scattered light direction) were the two significant parameters, and the diagnostic parameter HPS was established based on them. HPS showed a high sepsis diagnostic value, with an area under the ROC (AUC) of 0.862, a sensitivity of 70.0%, and a specificity of 87.1%. It is superior to the routine clinical parameters CRP (AUC 0.689, sensitivity 54.4%, and specificity 73.9%) and PCT (AUC 0.829, sensitivity 70.6%, and specificity 79.9%). The positive predictive value of HPS for the sepsis diagnosis was 70.3% in the validation cohort, which was similar to that in the training cohort at 66.0%. Conclusion: The HPS model improved the diagnostic performance of sepsis to some extent and could be used for early and rapid diagnosis of sepsis in the Emergency Department.

Sichuan Provincial People's Hospital Chengdu, China

Introduction: Multiple myeloma (MM) is the second most common blood cancer. Despite the improved survival rates in the era of modern treatment, MM remains invariably incurable, and most patients relapse. Multiple myeloma displays enormous heterogeneity, which represents a major obstacle to its diagnosis and treatment. The heterogeneous microenvironment of MM is crucial for myeloma development, treatment, and progression. Recent studies have highlighted the importance of monocytes in the microenvironment of MM not only for their number but also for their role in resistance to dexamethasone. However, the specific routes through which monocyte–plasma cell crosstalk accelerates disease progression have not yet been elucidated. We applied single-cell RNA sequencing to define newly diagnosed MM patients stratified by the Revised International Staging System (R-ISS) strategy to identify new potential pathways, therapeutic targets, and crosstalk with the microenvironment, particularly in the R-ISS stage III. Methods: We performed single-cell RNA sequencing of bone marrow samples obtained from two normal controls and nine MM patients to investigate the heterogeneity among the stages stratified by the R-ISS. Results: Plasma cells were divided into nine subgroups (P1 to P9), of which P1 to P5 existed only in stage III MM patients. PDIA6 was newly found to be enriched in P3 and involved in the unfolded protein response. In P1, LETM1 was upregulated and exhibited stem cell-like characteristics and high proliferation. There was also higher expression of PDIA6 and LETM1 observed in myloma cell lines and in myloma cells from 22 other MM patients than in plasma cells. At the single-cell level, we further explored whether the STAT1 regulon activated LETM1 via the C-type lectin receptor-signaling pathway and whether PDIA6 was putatively activated by the UQCRB regulon through oxidative phosphorylation to promote MM progression. Conclusions: This work investigated the expression profiles of myeloma cells and immune cells and identified putative markers and regulated networks in different R-ISS stages at the single-cell level, providing valuable insights into the heterogeneity of MM and into the therapeutic implications of having a comprehensive understanding of the genomic complexity of this disease.

¹Hematology – Università Cattolica del S. Cuore Rome Italy, ²Department of Cellular Biotechnologies and Haematology, "Sapienza" University of Rome Rome Italy, ³Fondazione Policlinico A. Gemelli IRCCS, Rome Italy

Introduction:. Digital morphology (DM) applied to hematology laboratories today represents a technological reality that facilitates the interpretation of instrumental alarms (ICSH guidelines) by proposing pre-classified cell images with better equipment standards (automated smearing, staining and a number of cells evaluated) than the manual ones. DM also enhances remote consultation, archiving, education, training and harmonization. Currently, there is a need for standardization of staining, magnification, color screen and file's format among available systems. The morphological skill remains fundamental for cell reclassification and diagnostic interpretation of DM, and it is undisputed that the correct use of DM in routine contributes positively to reducing false negatives, especially with samples containing pathological cells in low percentages. The realistic screen reproduction of blood cells as they appear under a light microscope is a crucial element for an effective, validated diagnostic screening in the routine DM pathway. Among the morphological alterations in peripheral blood, correctly identifying lymphocytes as normal, reactive and atypical is still very challenging. In two Universities in Rome, during a multicenter evaluation of Mindray MC80 system on peripheral blood (PB) samples containing qualitative or quantitative alterations, we evaluated the predictive ability of DM for lymphocyte lineage abnormalities. Methods:. Four PB smears from 223 patients with circulating abnormal or atypical lymphocytes have been anonymously collected at Sapienza University (Fig. 1): two for differential at optical microscopy (OM) on 200 cells (one for each center) to obtain the “Reference Differential” and two for analysis with the MC-80 in Catholic University to evaluate reproducibility and accuracy. Results:. The distribution range was 0.5%–99%, with excellent overall correlation OM versus DM with r² >0.86 and no false negative. Most of the disagreed samples were CLL with WBC count 33.9–195.6 × 10⁹/L in which several small-sized mature CD5+ lymphocytes were not classified as atypical lymphocytes. In these samples, marked lymphocytosis, number of smudge cells (Gümprecht shadows) and high reproducibility of chromatin pattern and cell size allow us to re-classify the cells on screen without resorting to OM. In six samples with normal WBC count (5.7–12.6 × 10⁹/L) and differential, rare circulating lymphoma cells were displayed on the screen with their unique morphological characteristics allowing the punctual and reproducible identification of the pathological sample (Fig. 1). Conclusions. Even in a demanding routine, the DM, in a context of morphological competence and reproducibility of morphological details, allows an excellent screening of pathological samples compared to the optical microscope.