Complex haemostatic abnormalities as a cause of bleeding after neurosurgery in a patient with Gaucher disease.

We report a treatment-naïve patient with Gaucher disease (GD) who experienced repeated bleeding after three neurosurgeries for a brain tumour, identified as an oligoastrocytoma. The patient had normal values on basic haemostatic tests: prothrombin time, 75-105%; activated partial thromboplastin time, 30.3-34 s; and mild thrombocytopaenia, 96-115 × 10(9 )cells/l. However, additional tests showed mild von Willebrand factor (vWF) deficiency (vWF antigen, 56%; vWF ristocetin cofactor, 49%; factor VIII [FVIII], 54%) and abnormal collagen-mediated platelet aggregation (0.45-0.55). Bleeding control was achieved after vWF/FVIII concentrate and platelet transfusions. This case raises questions about the safe platelet count and basic haemostatic tests for assessing bleeding risk in patients with GD prior to surgery. In patients with GD, a minimum haemostatic evaluation should include platelet count and basic haemostatic tests such as fibrinogen, prothrombin time, activated partial thromboplastin time as well as platelet function tests and assessing vWF and FVIII levels. Specific coagulation factors or platelet function deficiencies should be corrected with factor concentrates or platelet transfusions.

Acquired von Willebrand syndrome in patients with Gaucher disease.

Although various coagulation abnormalities occur in patients with Gaucher disease (GD), von Willebrand factor (vWF) deficiency has rarely been reported. A retrospective review of six treatment naïve cases with GD and concomitant vWF deficiency over a 12-year-period in a single center is presented. All patients had a personal history of prior hemorrhages. Based on both reduced level of vWF antigen (vWF:Ag, range 14-56%) and ristocetin cofactor activity (vWF:RCo, range 12-53%), with a vWF:RCo/Ag ratio >0.7, the diagnosis of type 1 von Willebrand disease was made in all six cases. During enzyme replacement therapy (ERT) of a 2-year duration all patients normalized their vWF:Ag levels. Based on the positive ERT effect on vWF:Ag levels, vWF deficiency was assumed to be acquired. It should be noted that beside vWF deficiency four patients with GD exhibited mild thrombocytopenia (range 81-131×10(9)/L) and three had additional hemostatic defects (reduced collagen platelet aggregation, FV, FXI and FXII deficiencies).

Comparison of two laboratory assays in monitoring the efficacy of different prophylaxis regimens for severe haemophilia.

BACKGROUND: The goal of this study was to compare the validity of two laboratory assays, rotation thromboelastometry (ROTEM) and endogenous thrombin potential (ETP), in monitoring and evaluating different prophylactic treatment regimens in patients with severe haemophilia. METHODS: Twenty adult patients with severe haemophilia were divided into three groups according to treatment regimen with concentrate of factor (F) VIII/IX: full-dose prophylaxis (5 patients), intermediate-dose prophylaxis (5 patients), and on demand treatment (10 patients). RESULTS: The ROTEM for the group treated with full-dose prophylaxis was significantly lower than ROTEM for the group treated with intermediate-dose prophylaxis (p = 0.025). Among the patients given full-dose prophylaxis, 40% (2 patients) had prolonged ROTEM after 3 months of treatment, while among those given intermediate-dose prophylaxis all patients (100%, 5 patients) had prolonged ROTEM (p = 0.038). The ETP was significantly improved after 3 months of full-dose in comparison with intermediate-dose prophylaxis (p = 0.042). CONCLUSIONS: ROTEM and ETP are useful laboratory assays for monitoring efficacy of different prophylaxis regimens with concentrate of FVIII/IX in patients with severe haemophilia, helping in making decisions regarding optimal dose-regimen prophylaxis.

The effect of paraprotein on platelet aggregation.

BACKGROUND: Some patients with paraproteinemia have platelet aggregation disorders and the aim of this study was to examine disturbance of platelet aggregation in healthy blood donors by isolated paraprotein in vitro. METHODS: Using Rivanol, paraprotein was separated from the serum of ten patients with paraproteinemia, who had decreased platelet aggregation with several inducers. Platelet aggregation in ten healthy donors was measured with and without addition of the isolated induced paraprotein. The test was repeated with added human immunoglobulins for intravenous use. RESULTS: Average of maximal levels of platelet aggregation has been significantly decreased in plasma rich in platelets (PRP) of healthy donors after addition of paraprotein when inducers are used: adenosine diphosphate (ADP) (P = 0.007), collagen (COL) (P = 0.008), ristocetin (RIS) (P = 0.001), and epinephrine (EPI) (P = 0.002). Average of latent time of platelet aggregation was significantly prolonged in healthy donors after addition of paraprotein with inducers: COL (P = 0.008), RIS (P = 0.008) and EPI (P = 0.006) while addition of human immunoglobulins caused no change in platelet aggregation. In comparison, when human immunoglobulins were added, maximal platelet aggregation and latent time did not change significantly. Paraprotein isolated from patients with paraproteinamia, who had decrease platelet aggregation, had significantly decreased platelet aggregation when added to PRP of healthy donors, in vitro. CONCLUSION: Platelet aggregation was not significantly changed was confirmed with addition of human immunoglobulins.

Immune tolerance induction in patients with severe hemophilia with inhibitors: expert panel views and recommendations for clinical practice.

For hemophilia patients with inhibitors, immune tolerance induction (ITI) may help to restore clinical response to factor (F) VIII or FIX concentrates. Several ITI regimens and protocols exist; however, despite 30 yr of progressive investigation, the ITI evidence base relies mainly on observational data. Expert opinion, experience, and interpretation of the available evidence are therefore valuable to support clinical decision-making. At the Sixth Zürich Haemophilia Forum, an expert panel considered recent data and consensus to distill key practice points relating to ITI. The panel supported current recommendations that, where feasible, ITI should be offered early to children and adults (ideally ≤ 5 yr of inhibitor detection) when inhibitor titers are <10 Bethesda units (BU) and should be stopped when successful tolerance is achieved. For hemophilia A inhibitor patients, ITI can be founded on recombinant FVIII at high doses. The panel considered that patients with a high bleeding frequency should be offered additional prophylaxis with a bypassing agent. For patients with hemophilia B, there may be a benefit of genetic testing to indicate the risk for inhibitors. ITI is often less effective and associated with a greater risk of side effects in these patients. For high-titer inhibitor (≥ 5 BU) hemophilia B patients, the panel advised that bypassing agents could be offered on demand in addition to ITI. Within future ITI regimens, there may be a role for additional immunosuppressant therapies. Participants agreed that research is needed to find alternatives to ITI therapy that offer durable and sustained effects and reduced rates of complications.

Haemostatic abnormalities in treatment-naïve patients with Type 1 Gaucher's disease.

There is a paucity of data on the effects of enzyme replacement therapy (ERT) on the coagulation abnormalities and platelet function of patients with Gaucher's disease (GDPs) and much of this data are controversial. This study investigates the haemostatic parameters in treatment-naïve GDPs and the effects of ERT. 31 Serbian treatment-naïve type 1 GDPs (M/F 17/14; median age 49 years, splenectomized 9/31) were studied. The complete blood count, prothrombin time (PT), activated partial tromboplastin time (aPTT) and coagulation factors were measured using the standard methods. Platelet aggregation was assessed with a whole-blood aggregometer. Splenic volumes were assessed using computer tomography. Twenty-one patients were treated with ERT (Imiglucerase). The haemostatic parameters were assessed after 6, 12 and 24 months (ERT(6, 12, 24)). Initially bleeding episodes were registered in 10/31 GDPs. Median platelet count was 108 × 10(9)/L; 22/31 GDPs were thrombocytopenic. The PT and aPTT values were abnormal in 16/31 and 13/31 GDPs, respectively. Platelet aggregation abnormalities were recorded in 19/31GDPs. Median platelet aggregation was reduced in response to adenosine-diphosphate 5 µmol/L (ADP(5) 0.46) and collagen 5 µmol/L (Col(5) 0.47). Splenic volume inversely correlated with the platelet count and a reduced response to arachidonic acid (AA), Col(5) and ADP(5) (p < 0.05). The splenectomized GDPs had a significantly lower platelet aggregation to Col(10) (p < 0.05). Bleeding GDPs had a significantly lower platelet count, higher chitotriosidase levels and a greater splenic volume compared to non-bleeding patients (p < 0.01). ERT: The number of bleeding GDPs had significantly decreased by ERT(6) (1/10; p < 0.01). The platelet count had significantly increased by ERT(6) (ERT(6) 180 × 10(9)/L, p < 0.01). The PT increased significantly from ERT(0) to ERT(24) (PT(0) 65%, PT(24) 81%; p < 0.05). The von Willebrand factor had increased significantly by ERT(6) and ERT(24) (ERT(0) 56%, ERT(6) 70%, ERT(12) 70%, ERT(24) 86%; p < 0.01). The number of GDPs with abnormal platelet aggregation had decreased significantly by ERT(6) (10/19; p < 0.05). Platelet aggregation on ADP(10) and AA significantly increased by ERT(6) (ADP(10): ERT(0) 0.75, ERT(6) 0.8 p < 0.01; AA: ERT(0) 0.7, ERT(6) 0.8 p < 0.05). In conclusion, platelet dysfunction and coagulation abnormalities were found in a considerable number of our GDPs. The absence of severe bleeding episodes suggests that the haemostatic system is sufficiently balanced and therefore the exact mechanism of the etiology of these abnormalities need to be fully clarified. ERT resulted in the cessation of bleeding and marked increase in platelet count, PT, vWF and platelet aggregation.

Veno-occlusive disease in pediatric patients after hematopoietic stem cell transplantation: relevance of activated coagulation and fibrinolysis markers and natural anticoagulants.

Prediction of veno-occlusive disease (VOD), its precise diagnosis, and treatment have been the subject of various studies, but still remain unclear. Our goal was to investigate the levels of activated coagulation and fibrinolysis markers and natural anticoagulants in pediatric patients with VOD after hematopoietic stem cell transplantation (HSCT). We investigated 47 pediatric patients: 20 with neuroblastoma, 17 with leukemias, and 10 with lymphomas and measured the values of antithrombin (AT), protein C (PC), fibrinogen (FI), thrombin AT complex, prothrombin fragments 1+2 (F1+2), and D-dimer from day -7 to day +30 post-HSCT. Patients were monitored for the occurrence of VOD, and it occurred in 10 patients at a median post-HSCT day of 17.5 (range: 2 to 28 d). In the VOD group, at baseline the levels of FI were significantly lower, and on days +7 and +14 a relevant difference existed in F1+2 levels. The levels of PC were significantly lower on day +14. Logistic multivariate regression analysis between the groups showed significantly different D-dimer levels on day +14. On day +30, the levels of PC, AT, and F1+2 were different between these 2 groups of patients. The levels of D-dimer and F1+2 were increased, and PC and FI decreased before the clinical onset of VOD. The parameter differences may have a predictive value in VOD onset, which makes them candidates to be routinely monitored in patients after HSCT.

Correlation to FVIII:C in Two Thrombin Generation Tests: TGA-CAT and INNOVANCE ETP.

INTRODUCTION: Several thrombin-generation tests are available, but few have been directly compared. Our primary aim was to investigate the correlation of two thrombin generation tests, thrombin generation assay-calibrated automated thrombogram (TGA-CAT) and INNOVANCE ETP, to factor VIII levels (FVIII:C) in a group of patients with hemophilia A. The secondary aim was to investigate inter-laboratory variation for the TGA-CAT method. METHODS: Blood samples were taken from 45 patients with mild, moderate and severe hemophilia A. The TGA-CAT method was performed at both centers while the INNOVANCE ETP was only performed at the Stockholm center. Correlation between parameters was evaluated using Spearman's rank correlation test. For determination of the TGA-CAT inter-laboratory variability, Bland-Altman plots were used. RESULTS: The correlation for the INNOVANCE ETP and TGA-CAT methods with FVIII:C in persons with hemophilia (PWH) was r=0.701 and r=0.734 respectively.The correlation between the two methods was r=0.546.When dividing the study material into disease severity groups (mild, moderate and severe) based on FVIII levels, both methods fail to discriminate between them.The variability of the TGA-CAT results performed at the two centers was reduced after normalization; before normalization, 29% of values showed less than ±10% difference while after normalization the number increased to 41%. CONCLUSIONS: Both methods correlate in an equal manner to FVIII:C in PWH but show a poor correlation with each other. The level of agreement for the TGA-CAT method was poor though slightly improved after normalization of data. Further improvement of standardization of these methods is warranted.

Iliopsoas haematoma in Gaucher disease.

Extrapelvis or retroperitoneal haemorrhage has long been appreciated as having many causes and considerable variability in subsequent morbidity; however, to date only two cases have been reported in patients with Gaucher disease, the most common lysosomal storage disorder. It had been our assumption that these cases were unique and a consequence of severe disease in patients who had not been treated with enzyme (or other disease-specific) therapy. Herein we present three more cases (as well as our first patient), which allow one to make some generalizations. Ultrasound was used in one centre and computed tomography in the second centre to make the definitive diagnosis. The trigger for the bleeding in all cases was muscle strain after activity. All patients were young with massive hepatosplenomegaly, anaemia, thrombocytopenia, and bone pain with skeletal involvement; the last was the most obvious commonality among these patients. Differential diagnosis is complicated by exquisite groin pain that is common to both Gaucher disease and extrapelvis haemorrhage, but not necessarily.

Myeloid sarcoma of the skin in a patient with myelodysplastic syndrome.

We report the case of a 76-year-old woman who presented with asymptomatic extensive erythematous. Firm plaques were noted over the right cheek. Complete blood count was normal, as was a peripheral smear. An excision biopsy taken from the cheek showed infiltration of the dermis and hypodermis with atypical cells which were strongly positive for human leukocyte antigen (HLA-DR) and lysozyme and were moderately myeloperoxidase (MPO) enzyme. The results of immunohistochemical staining for CD34, CD117, CD3, CD4, CD8, CD20, CD23, CD56, and ALK-1 were negative. Bone marrow analysis indicated myelodysplastic syndrome RAEB 1 while cytogenetic finding showed tetrasomy 8. It was recommended that the patient undergo local radiotherapy of skin lesions, but she refused and was lost to follow-up.

Diagnostic challenges during pretreatment long-term follow-up in a patient with FIP1L1-PDGFRA-positive eosinophilia.

Obtaining a precise characterization of eosinophilia is crucial, as successful treatment relies on the underlying etiology of the disease. Platelet-derived growth factor receptor alpha-related disorders were first specified in 2008 as a distinct group of clonal eosinophilic disorders with exceptional responsiveness to imatinib. We herein present the case of a man with myeloid neoplasm and eosinophilia in whom a definitive diagnosis could not be adequately made based on histopathological features who was ultimately diagnosed only after extensive molecular analyses and successfully treated with imatinib. In addition, we discuss the diagnostic and therapeutic approaches to treating patients presenting with eosinophilia.

Rationale for individualizing haemophilia care.

Owing to the heterogeneity in the clinical phenotype of haemophilia A and B, it is now recognized that disease severity (based on factor VIII/IX activity) may no longer be the most appropriate guide for treatment and that a 'one-size-fits-all' approach is unlikely to achieve optimal therapy. Based on the present literature and consensus views of a group of experts in the field, this article highlights key gaps in the understanding of the diverse relationships between bleeding phenotype and factors such as joint health, genetic susceptibility, laboratory parameters, quality of life and management of pain. Early prophylaxis is a potential 'gold standard' therapy and issues surrounding inhibitor development, variations in its clinical use and long-term outcomes are discussed. Comprehensive treatment should be individualized for all patients (including those with mild or moderate haemophilia and carriers). Wherever possible all patients should be given prophylaxis. However, adult patients with a milder haemophilia phenotype may be candidates for ceasing prophylaxis and switching to on-demand treatment. Regardless, all treatment (on-demand and prophylaxis) should be tailored towards both the patient's personal needs and their clinical profile. In addition, as the associations between risk factors (psychosocial, condition-related and treatment-related) and clinical features are unique to each patient, an individualized approach is required to enable patients to alter their behaviour in response to them. The practical methodologies needed to reach this goal of individualized haemophilia care, and the health economic implications of this strategy, are ongoing topics for discussion.

Comparison of standard fibrinogen measurement methods with fibrin clot firmness assessed by thromboelastometry in patients with cirrhosis.

BACKGROUND: The Clauss fibrinogen method and thrombin clotting time (TCT) are still routinely used in patients with cirrhosis to define fibrinogen concentration and clotting potential. The thromboelastometric functional fibrinogen FIBTEM assay evaluates the strength of fibrin-based clots in whole blood, providing information on both quantitative deficit and fibrin polymerization disorders. OBJECTIVE: To compare these three methods of assessing fibrinogen in patients with cirrhosis of different aetiologies, characterized by impairment in fibrinogen concentration as well as functional aberrance. METHODS: Sixty patients with alcoholic and 24 patients with cholestatic cirrhosis were included (Child-Pugh score (CPs)A, n=24; B, n=32; C, n=28). All parameters were compared with those from a control group. Maximum clot firmness (MCF) in the FIBTEM test was assessed in regard to its relevance in detection of qualitative fibrinogen disorders in comparison with results obtained by standard measurement methods, i.e. the Clauss fibrinogen method and TCT. RESULTS: With increased cirrhosis severity, fibrinogen and FIBTEM-MCF levels significantly declined (p=0.002), while TCT was significantly prolonged (p=0.002). In all CPs groups, fibrinogen strongly correlated with FIBTEM-MCF (r=0.77, r=0.72, r=0.74; p<0.001), while cross-correlations of other assays were highly variable. The prevalence of decreased FIBTEM-MCF values (<9 mm) was significantly higher in advanced CPs categories (p=0.027), whereby the highest prevalence was detected in patients with CPsC (10/16; 62.5%). Nine of the 16 patients with decreased FIBTEM-MCF values had also decreased fibrinogen levels, while in the remaining 7 patients fibrinogen levels were within the reference range, indicating the possible presence of qualitatively altered fibrinogen that could be detected by FIBTEM-MCF. CONCLUSIONS: FIBTEM-MCF may be considered as a reliable alternative to standard plasma fibrinogen measurement in cirrhotic patients, especially in evaluating fibrin polymerization disorders in these patients. Further studies are needed to evaluate the usefulness of this assay in predicting bleeding complications in cirrhotic patients as well as monitoring replacement treatment.

Thrombotic events in acute promyelocytic leukemia.

INTRODUCTION: Thrombotic events (TE) appear to be more common in acute promyelocytic leukemia (APL) than in other acute leukemias, with reported prevalence ranging from 2 to 10-15%. MATERIALS AND METHODS: We retrospectively analyzed the data on TE appearance in 63 APL patients. RESULTS: TE occured in 13 (20.6%) cases, four arterial (6.3%) and nine venous (14.3%). TE were more frequently diagnosed after initiation of weekly D-dimer monitoring (7 TE during 20 months vs 6 during 76 months, P=0.032). Patients with and without venous thrombosis were significantly different regarding female/male ratio (P=0.046), PT (P=0.022), aPTT (P=0.044), ISTH DIC score (P=0.001), bcr3 (P=0.02) and FLT3-ITD (P=0.028) mutation. The most significant risk factor for venous TE occurrence in multivariate analysis was FLT3-ITD mutation (P=0.034). PAI-1 4G/4G polymorphism was five times more frequent in patients with venous TE than without it (P=0.05). Regarding risk factors for arterial TE we failed to identify any. CONCLUSIONS: We have demonstrated that APL-related TE rate is higher than previously reported and that weekly D-dimer monitoring might help to identify patients with silent thrombosis. Moreover, our study suggests a possible relationship between venous TE occurrence and several laboratory findings (PT, aPTT, ISTH DIC score, bcr3 isoform, FLT3-ITD mutation and PAI 4G/4G). Prophylactic use of heparin might be considered in patients with ISTH DIC score<5, bcr3 isoform, FLT3-ITD mutation and PAI 4G/4G.

111In-platelets dynamic study in chronic immune thrombocytopenic purpura.

BACKGROUND: The aim of this work was to estimate the significance of a dynamic study performed during the first 20 minutes after autologous (111)In-oxinate-platelets injection in patients with chronic immune thrombocytopenic purpura (ITP). Two hypotheses were tested: a) dynamic study indicates the place of platelet sequestration; b) dynamic study reflects the quality of platelet separation and labelling procedure. MATERIAL AND METHODS: Thirty-nine persons were investigated: 25 with shortened platelet life span (ITP), and 14 with normal platelet life span (6 healthy subjects and eight patients with myelodysplastic syndrome--MDS). Platelet blood count on the day of platelet labelling, general yield of platelet labelling (GYL), differential yield of platelet labelling (DYL), platelet life span, dynamic study with initial platelet accumulation in the liver (IPAL), sequential static study for determining the platelet sequestration index (SI) and platelet sequestration site (SS) were investigated. RESULTS: Two types of labelled platelet kinetics were determined in both groups of patients: IPAL < 20% and IPAL > 20%. A statistically significant difference in GYL and DYL was noted between the patients with IPAL < 20% and IPAL > 20%. No significant difference was registered in platelet blood count, life span, SS and SI between the two groups of patients. Both yields of platelet labelling were higher in the group with IPAL < 20%. There was no correlation between IPAL and platelet SI, or between IPAL and platelet SS. CONCLUSIONS: Dynamic study with (111)In-platelets cannot predict platelet sequestration site in ITP patients, but it is a useful and sensitive method of platelet labelling procedure quality control.

Clinical significance of optimal red cell mass and plasma volume estimation methods.

BACKGROUND: The aim of this study was to present and compare the results of proposed methods for optimal red cell mass and plasma volume (RCM&PV) estimation, and their influence on the interpretation of obtained results. MATERIAL AND METHODS: In 120/280 patients with polycythaemia rubra vera, subjected to RCM&PV determination with autologous erythrocytes in vitro labelled with 51Cr-sodium chromate, optimal volumes were determined using: 1. traditional ml/kg of: --the real body weight method (ml/kg RBW); --the optimal body weight method (ml/kg OBW). 2. the body weight, height, and sex based method (Retzlaff's tables), 3. the method recommended by the International Council for Standardization in Haematology (ICSH), based on body surface area. RESULTS: Different interpretation of the same results of 120 RCM&PV measurements was registered in 48/120 patients (40%). The greatest disagreement existed between ml/kg RBW and ml/kg OBW methods (in 39/120 subjects, 32.5%). In underweight patients the ml/kg RBW method, and in overweight patients the ml/kg OBW method, offered better agreement with ICSH&Retzlaff's methods. The ml/kg RBW method disagreed with ICSH&Retzlaff's methods and ml/kg OBW in 25% and 19.2% of patients respectively. ICSH and Retzlaff's methods disagreed in 10/120 patients (8.3%). The ICSH method yielded significantly lower optimal volumes than Retzlaff's. CONCLUSION: Three methods for optimal RCM&PV estimation lead to different interpretations of the same results of RCM&PV measurements with 51Cr-erythrocytes in 40% of patients. Two ml/kg body weight methods show greater disagreement in comparison with ICSH and Retzlaff's methods, which differ significantly. The ICSH method yields lower optimal values compared to Retzlaff's.

[Obstetric and gynecological intervention in women with Bernard-Soulier syndrome: report of two cases].

Introduction Bernard-Soulier syndrome (BSS) is a rare inherited bleeding disorder characterized by giant platelets thrombocytopenia, prolonged bleeding time, frequent hemorrhages with considerable morbidity. Data on the outcome of pregnancy and gynecological intervention in BSS are rare and there are no general therapeutic recommendations. CASES OUTLINE: We report two cases of BSS. In the first case a 29-year-old patient with BSS was admitted in 8 weeks of gestation. The diagnosis of BSS was made on the basis of prolonged bleeding time, giant-platelets thrombocytopenia, and absent ristocetin-induced platelet aggregation. In 38 week of gestation Cesarean section, with platelets transfusion preparation, was performed. Obstetric intervention passed without complication. Postoperative course was complicated with a three-week vaginal bleeding resistant to platelet transfusion. Neonate platelet count was normal. Our second case was a 28-year-old patient with BSS, hospitalized for ovarial tumor surgery. The patient was prepared for the intervention with platelets transfusion. The surgery was uncomplicated, but on the second postoperative day a massive vaginal bleeding, resistant to the platelet transfusion, developed. Bleeding control was achieved with activated recombinant factor VII. Twelve hours the patient developed later hypertensive crisis with epileptic seizure due to subarachnoid hemorrhage. Therapy was continued with platelet transfusion, antihypertensive and antiedema drugs. PH examination of tumor tissue showed hemorrhagic ovarial cyst. CONCLUSION: Obstretic and gynecological intervention in women with BSS may be associated with a life-threatening bleeding thus requiring a multidisciplinary approach with adequate preparation. Because of the limited data in the literature, it is not possible to provide firm management recommendations and each case should be managed individually.

The importance of cardiovascular risk factors for thrombosis prediction in patients with essential thrombocythemia.

The current widely accepted stratification defined by age and previous thrombosis in patients with essential thrombocythemia (ET) probably deserves deeper analysis. The aim of our study was to identify additional factors at the time of diagnosis, which have an impact on the thrombosis prediction. We conducted a study of 244 consecutive ET patients with median follow-up of 83 months. We analyzed the influence of age, gender, laboratory parameters, history of previous thrombosis, spleen size, JAK2 mutation as well as cardiovascular (CV) risk factors including arterial hypertension, diabetes, active tobacco use and hyperlipidemia in the terms of thrombosis. The most important predictors of thrombosis in multivariate Cox regression model were the presence of CV risk factors (p=0.004) and previous thrombosis (p=0.038). Accordingly, we assigned risk scores based on multivariable analysis-derived hazard ratios (HR) to the presence of 1 CV risk factor (HR=3.5; 1 point), >1 CV risk factors (HR=8.3; 2 points) and previous thrombosis (HR=2.0; 1 point). A final three-tiered prognostic model for thrombosis prediction was developed as low (score 0), intermediate (score 1 or 2) and high risk (score 3) (p<0.001). The hazard of thrombosis was 3.8% in low-risk group, 16.7% in the intermediate-risk group and 60% in the high-risk group (p<0.001). Patients with thrombotic complications during the follow-up had a significantly shorter survival (p=0.018). The new score based on CV risk factors and previous thrombotic events allows a better patient selection within prognostic-risk groups and improved identification of the high-risk patients for thrombosis.