Search

BACKGROUND Diagnosing heparin-induced thrombocytopenia (HIT) at the bedside remains challenging, exposing a significant number of patients at risk of delayed diagnosis or overtreatment. We hypothesized that machine-learning algorithms could be utilized to develop a more accurate and user-friendly diagnostic tool that integrates diverse clinical and laboratory information and accounts for complex interactions. METHODS We conducted a prospective cohort study including 1393 patients with suspected HIT between 2018 and 2021 from 10 study centers. Detailed clinical information and laboratory data were collected, and various immunoassays were conducted. The washed platelet heparin-induced platelet activation assay (HIPA) served as the reference standard. FINDINGS HIPA diagnosed HIT in 119 patients (prevalence 8.5%). The feature selection process in the training dataset (75% of patients) yielded the following predictor variables: (1) immunoassay test result, (2) platelet nadir, (3) unfractionated heparin use, (4) CRP, (5) timing of thrombocytopenia, and (6) other causes of thrombocytopenia. The best performing models were a support vector machine in case of the chemiluminescent immunoassay (CLIA) and the ELISA, as well as a gradient boosting machine in particle-gel immunoassay (PaGIA). In the validation dataset (25% of patients), the AUROC of all models was 0.99 (95% CI: 0.97, 1.00). Compared to the currently recommended diagnostic algorithm (4Ts score, immunoassay), the numbers of false-negative patients were reduced from 12 to 6 (-50.0%; ELISA), 9 to 3 (-66.7%, PaGIA) and 14 to 5 (-64.3%; CLIA). The numbers of false-positive individuals were reduced from 87 to 61 (-29.8%; ELISA), 200 to 63 (-68.5%; PaGIA) and increased from 50 to 63 (+29.0%) for the CLIA. INTERPRETATION Our user-friendly machine-learning algorithm for the diagnosis of HIT (https://toradi-hit.org) was substantially more accurate than the currently recommended diagnostic algorithm. It has the potential to reduce delayed diagnosis and overtreatment in clinical practice. Future studies shall validate this model in wider settings. FUNDING Swiss National Science Foundation (SNSF), and International Society on Thrombosis and Haemostasis (ISTH).

Prothrombinase-induced clotting time (PiCT) is proposed as a rapid and inexpensive laboratory test to measure direct oral anticoagulant (DOAC) drug levels. In a prospective, multicenter cross-sectional study, including 851 patients, we aimed to study the accuracy of PiCT in determining rivaroxaban, apixaban, and edoxaban drug concentrations and assessed whether clinically relevant drug levels could be predicted correctly. Citrated plasma samples were collected, and the Pefakit® PiCT was utilized. Ultra-high performance liquid chromatography-tandem mass spectrometry (LC-MS/MS) was performed to measure drug concentrations. Cut-off levels were established using receiver-operating characteristics curves. We calculated sensitivities and specificities with respect to clinically relevant drug concentrations. Spearman's correlation coefficient between PiCT and drug concentrations was 0.85 in the case of rivaroxaban (95% CI 0.82, 0.88), 0.66 for apixaban (95% CI 0.60, 0.71), and 0.78 for edoxaban (95% CI 0.65, 0.86). The sensitivity to detect clinically relevant drug concentrations was 85.1% in the case of 30 µg L-1 (95% CI 82.0, 87.7; specificity 77.9; 72.1, 82.7), 85.7% in the case of 50 µg L-1 (82.4, 88.4; specificity 77.3; 72.5, 81.5), and 85.1% in the case of 100 µg L-1 (80.9, 88.4; specificity 73.2%; 69.1, 76.9). In conclusion, the association of PiCT with DOAC concentrations was fair, and the majority of clinically relevant drug concentrations were correctly predicted.

Platelets interact with multiple adhesion proteins during thrombogenesis, yet little is known about their ability to assemble fibronectin matrix. In vitro three-dimensional superresolution microscopy complemented by biophysical and biochemical methods revealed fundamental insights into how platelet contractility drives fibronectin fibrillogenesis. Platelets adhering to thrombus proteins (fibronectin and fibrin) versus basement membrane components (laminin and collagen IV) pull fibronectin fibrils along their apical membrane versus underneath their basal membrane, respectively. In contrast to other cell types, platelets assemble fibronectin nanofibrils using αIIbβ3 rather than α5β1 integrins. Apical fibrillogenesis correlated with a stronger activation of integrin-linked kinase, higher platelet traction forces, and a larger tension in fibrillar-like adhesions compared to basal fibrillogenesis. Our findings have potential implications for how mechanical thrombus integrity might be maintained during remodeling and vascular repair., Science Advances, 8 (10), ISSN:2375-2548

Applying a single anti-Xa assay, calibrated to unfractionated heparin to measure rivaroxaban, apixaban, and edoxaban would simplify laboratory procedures and save healthcare costs. We hypothesized that a heparin-calibrated anti-Xa assay would accurately measure rivaroxaban, apixaban, and edoxaban drug concentrations and correctly predict clinically relevant drug levels. This analysis is part of the Simple-Xa study, a prospective multicenter cross-sectional study conducted in clinical practice. Patients treated with rivaroxaban, apixaban, or edoxaban were included. Anti-Xa activity was measured using the Siemens INNOVANCE ® Heparin assay. Drug concentrations were determined using ultra-high performance liquid chromatography-tandem mass spectrometry (LC-MS/MS). Cut-off levels were determined in a derivation dataset (50% of patients) and sensitivities and specificities were calculated in a verification dataset (50% of patients). Overall, 845 patients were available for analysis. Correlation coefficients (r s ) between the heparin-calibrated anti-Xa assay and drug concentrations were 0.97 (95% CI 0.97, 0.98) for rivaroxaban, 0.96 (0.96, 0.97) for apixaban, and 0.96 (0.94, 0.99) for edoxaban. The area under the receiver operating characteristics curve (ROC) was 0.99 for all clinically relevant drug concentrations. In the verification dataset, the sensitivity was 94.2% (95% CI 90.8-96.6) for 30 μg L -1 , 95.8% (92.4-98.0) for 50 μg L -1 , and 98.7% (95.5-99.9) for 100 μg L -1 . Specificities were 86.3% (79.2-91.7), 89.8% (84.5-93.7), and 88.7% (84.2-92.2), respectively. In a large prospective study in clinical practice, a strong correlation of heparin-calibrated anti-Xa measurements with LC-MS/MS results was observed and clinically relevant drug concentrations were predicted correctly.

Background The differential diagnosis of abdominal pain in pregnant women is broad. Liver diseases as the origin of abdominal pain in pregnancy are rare, and severe forms occur in less than 0.1% of pregnancies. Some disorders, such as hemolysis, elevated liver enzymes, low platelets (HELLP) syndrome and preeclampsia, are unique to pregnancy, while others, such as antiphospholipid antibody syndrome, may manifest in pregnancy but have consequences beyond the current pregnancy. All of them require prompt identification and treatment. Case presentation A 27-year-old Caucasian woman who was 15+1 weeks pregnant reported to the emergency department twice due to stabbing right-upper-quadrant abdominal pain. Initial laboratory testing revealed mild leukocytosis and slightly elevated liver enzymes. On second presentation, the patient was febrile and had an increased C-reactive protein concentration. Over the course of the next days, nonhemolytic anemia and thrombocytopenia emerged with elevated liver enzymes. Coagulation studies also revealed a prolongation of activated partial thromboplastin time. Magnetic resonance imaging showed nonspecific alterations in the right liver lobe, possibly corresponding to infection or infarction. A hepatic viral infection was ruled out. At that time, the most likely diagnosis was cholangitis with liver abscess formation, and antibiotic therapy was started. Further worsening of the anemia and thrombocytopenia, development of proteinuria, together with a miscarriage on the fourth day of hospitalization resulted in the tentative diagnosis of (triple-positive) antiphospholipid antibody syndrome, which was confirmed 12 weeks after the initial investigation. Treatment consisted of prompt anticoagulation with heparin and later on with a vitamin K antagonist as well as high-dose glucocorticoid therapy. There was no need for intravenous immunoglobulin therapy or plasma exchange, although we suspected a catastrophic form of antiphospholipid antibody syndrome due to infarctions of the liver, placenta, and possibly kidneys (proteinuria). The outcome was favorable. Conclusion We report a 27-year-old pregnant woman whose abdominal pain was caused by liver infarctions as the first manifestation of catastrophic antiphospholipid antibody syndrome. The antiphospholipid antibody syndrome was possibly secondary to hitherto clinically silent systemic lupus erythematosus since the antinuclear antibodies were increased later on. Hydroxychloroquine therapy was initiated to prevent antiphospholipid antibody syndrome recurrence in a future pregnancy.

Background: The thrombin generation assay (TG) is a promising approach to measure the degree of anticoagulation in patients treated with direct oral anticoagulants (DOAC). A strong association with plasma drug concentrations would be a meaningful argument for the potential use to monitor DOAC. Objectives: We aimed to study the correlation of TG with rivaroxaban, apixaban, and edoxaban drug concentrations in a large, prospective multicenter cross-sectional study. Methods: Five-hundred and fifty-nine patients were included in nine tertiary hospitals. The Technothrombin® TG was conducted in addition to an anti-Xa assay; LC-MS/MS was performed as the reference standard. Results: Correlation (r s ) between thrombin generation measurements and drug concentrations was -0.72 for peak thrombin generation (95% confidence interval, CI, -0.77, -0.66), -0.55 for area under the curve (AUC; 95% CI -0.61, -0.48), and 0.80 for lag time (95% CI 0.75, 0.84). In contrast, r s was 0.96 with results of the anti-Xa activity (95% CI 0.95-0.97). Sensitivity with regard to the clinically relevant cut-off value of 50 μgL -1 was 49% in case of peak thrombin generation (95% CI, 44, 55), 29% in case of AUC (95% CI, 24, 34), and 64% in case of lag time (95% CI, 58, 69). Sensitivity of the anti-Xa assay was 95% (95% CI, 92, 97). Conclusions: The correlation of thrombin generation measurements with DOAC drug concentrations was weak, and clinically relevant drug levels were not predicted correctly. Our results do not support an application of TG in the monitoring of DOAC.