Search

This work focusses on extending the deep compartment model (DCM) framework to the estimation of mixed-effects. By introducing random effects, model predictions can be personalized based on drug measurements, enabling the testing of different treatment schedules on an individual basis. The performance of classical first-order (FO and FOCE) and machine learning based variational inference (VI) algorithms were compared in a simulation study. In VI, posterior distributions of the random variables are approximated using variational distributions whose parameters can be directly optimized. We found that variational approximations estimated using the path derivative gradient estimator version of VI were highly accurate. Models fit on the simulated data set using the FO and VI objective functions gave similar results, with accurate predictions of both the population parameters and covariate effects. Contrastingly, models fit using FOCE depicted erratic behaviour during optimization, and resulting parameter estimates were inaccurate. Finally, we compared the performance of the methods on two real-world data sets of haemophilia A patients who received standard half-life factor VIII concentrates during prophylactic and perioperative settings. Again, models fit using FO and VI depicted similar results, although some models fit using FO presented divergent results. Again, models fit using FOCE were unstable. In conclusion, we show that mixed-effects estimation using the DCM is feasible. VI performs conditional estimation, which might lead to more accurate results in more complex models compared to the FO method. [ABSTRACT FROM AUTHOR]

In population pharmacokinetic (PK) models, interindividual variability is explained by implementation of covariates in the model. The widely used forward stepwise selection method is sensitive to bias, which may lead to an incorrect inclusion of covariates. Alternatives, such as the full fixed effects model, reduce this bias but are dependent on the chosen implementation of each covariate. As the correct functional forms are unknown, this may still lead to an inaccurate selection of covariates. Machine learning (ML) techniques can potentially be used to learn the optimal functional forms for implementing covariates directly from data. A recent study suggested that using ML resulted in an improved selection of influential covariates. However, how do we select the appropriate functional form for including these covariates? In this work, we use SHapley Additive exPlanations (SHAP) to infer the relationship between covariates and PK parameters from ML models. As a case‐study, we use data from 119 patients with hemophilia A receiving clotting factor VIII concentrate peri‐operatively. We fit both a random forest and a XGBoost model to predict empirical Bayes estimated clearance and central volume from a base nonlinear mixed effects model. Next, we show that SHAP reveals covariate relationships which match previous findings. In addition, we can reveal subtle effects arising from combinations of covariates difficult to obtain using other methods of covariate analysis. We conclude that the proposed method can be used to extend ML‐based covariate selection, and holds potential as a complete full model alternative to classical covariate analyses. [ABSTRACT FROM AUTHOR]

Nonlinear mixed effect (NLME) models are the gold standard for the analysis of patient response following drug exposure. However, these types of models are complex and time‐consuming to develop. There is great interest in the adoption of machine‐learning methods, but most implementations cannot be reliably extrapolated to treatment strategies outside of the training data. In order to solve this problem, we propose the deep compartment model (DCM), a combination of neural networks and ordinary differential equations. Using simulated datasets of different sizes, we show that our model remains accurate when training on small data sets. Furthermore, using a real‐world data set of patients with hemophilia A receiving factor VIII concentrate while undergoing surgery, we show that our model more accurately predicts a priori drug concentrations compared to a previous NLME model. In addition, we show that our model correctly describes the changing drug concentration over time. By adopting pharmacokinetic principles, the DCM allows for simulation of different treatment strategies and enables therapeutic drug monitoring. [ABSTRACT FROM AUTHOR]

Summary: Congenital platelet disorders (CPDs) are rare bleeding disorders that are associated with mucocutaneous bleeds. However, data on vaginal bleeding in women with CPDs are scarce. A set of generic and bleeding‐specific questionnaires were used to evaluate the prevalence of vaginal bleeding, its impact on quality of life (QoL) and sexual functioning and the consequences for pregnancy, miscarriage and delivery in a cohort of women who were referred for diagnostic evaluation for CPDs. A total of 78 women included in the study were either diagnosed with a CPD (n = 35) or were clinically suspected of a CPD (n = 43). Heavy menstrual bleeding (HMB) was reported by a large proportion of women, which mainly started at menarche. In all, 76% of women received any kind of HMB treatment, often leading to surgical prodecures. HMB was shown to have a high impact on QoL, which improved upon treatment. Even though women reported that vaginal bleeding affects sexuality, this topic is not frequently discussed with physicians. Heavy blood loss frequently occurred after miscarriage/delivery, often requiring treatment. Women with (suspected) CPDs frequently encounter HMB, negatively impacting daily life and sexual functioning. Together with peripartum bleeding, these data highlight the burden of vaginal bleeding in CPDs and importance of adequate treatment. [ABSTRACT FROM AUTHOR]

OBJECTIVE Primary aim: to determine the prevalence of bleedings and thrombotic events in anticoagulants users during hospitalization. Secondary aims: to determine the type and location and the potential risk factors that are associated with bleedings and thrombotic events. DESIGN Prospective, observational, multicenter cohort study. METHODS The study was performed at the Erasmus MC, Rotterdam and the Reinier de Graaf hospital, Delft. Patients who were treated with anticoagulants were included in the study. Exclusion criteria were hospitalization of less than 24 hours, admission to the intensive care unit without admission to another, regular hospital department, and patients treated with low-molecular-weight heparins only for thrombosis prophylaxis. RESULTS We included 942 patients. The prevalence of in-hospital bleeding events was 8.8%, of which 44% were surgical site bleedings. The prevalence of thrombotic events during hospitalization was 0.7%. Multivariate logistic regression analysis indicated that female gender (odds ratio [OR] 1.82; 95% confidence interval [95% CI] 1.11-2.99), age 70 years or older (OR 1.73; 95% CI 1.03-2.89), hospitalization of five days or longer (OR 2.86; 95% CI 1.43-5.72), admission to the academic hospital (OR 1.93; 95% CI 1.12-3.34) and admission to surgical wards (OR 1.98; 95% CI 1.20-3.73) were associated with more bleeding events. We found no potential risk factors associated with thrombotic events, as the number of cases was very small. CONCLUSION Bleeding events occurred in 8.8%, thrombotic events in 0.7% of anticoagulants users during hospitalization. Female gender, age 70 years or older, hospitalization of 5 days or longer, admission to the academic hospital and admission to surgical wards seem to be risk factors. [ABSTRACT FROM AUTHOR]

BACKGROUND Pulmonary embolism is one of the leading causes of maternal death in the Western world. Because of the low specificity and sensitivity of the D-dimer test, all pregnant women with suspected pulmonary embolism undergo computed tomographic (CT) pulmonary angiography or ventilation-perfusion scanning, both of which involve radiation exposure to the mother and fetus. Whether a pregnancy-adapted algorithm could be used to safely avoid diagnostic imaging in pregnant women with suspected pulmonary embolism is unknown. METHODS In a prospective study involving pregnant women with suspected pulmonary embolism, we assessed three criteria from the YEARS algorithm (clinical signs of deep-vein thrombosis, hemoptysis, and pulmonary embolism as the most likely diagnosis) and measured the D-dimer level. Pulmonary embolism was ruled out if none of the three criteria were met and the D-dimer level was less than 1000 ng per milliliter or if one or more of the three criteria were met and the D-dimer level was less than 500 ng per milliliter. Adaptation of the YEARS algorithm for pregnant women involved compression ultrasonography for women with symptoms of deep-vein thrombosis; if the results were positive (i.e., a clot was present), CT pulmonary angiography was not performed. All patients in whom pulmonary embolism had not been ruled out underwent CT pulmonary angiography. The primary outcome was the incidence of venous thromboembolism at 3 months. The secondary outcome was the proportion of patients in whom CT pulmonary angiography was not indicated to safely rule out pulmonary embolism. RESULTS A total of 510 women were screened, of whom 12 (2.4%) were excluded. Pulmonary embolism was diagnosed in 20 patients (4.0%) at baseline. During follow-up, popliteal deep-vein thrombosis was diagnosed in 1 patient (0.21%; 95% confidence interval [Cl], 0.04 to 1.2); no patient had pulmonary embolism. CT pulmonary angiography was not indicated, and thus was avoided, in 195 patients (39%; 95% Cl, 35 to 44). The efficiency of the algorithm was highest during the first trimester of pregnancy and lowest during the third trimester; CT pulmonary angiography was avoided in 65% of patients who began the study in the first trimester and in 32% who began the study in the third trimester. CONCLUSIONS Pulmonary embolism was safely ruled out by the pregnancy-adapted YEARS diagnostic algorithm across all trimesters of pregnancy. CT pulmonary angiography was avoided in 32 to 65% of patients. (Funded by Leiden University Medical Center and 17 other participating hospitals; Artemis Netherlands Trial Register number, NL5726.) [ABSTRACT FROM AUTHOR]

A 28-year-old patient with severe haemophilia A presented to the emergency department with significant and painful swelling of the left cheek, an extensive haematoma extending from the left ear to the anterior thoracic region, an intraoral haematoma over the soft palate with deviation of the uvula to the right, and complaints of shortness of breath when lying down. Three days prior, his dentist had performed a restoration of the 36 molar under local anaesthesia. Due to pain, the general practitioner had administered an intramuscular injection of a non-steroidal anti-inflammatory drug (NSAID) two days post-procedure. The patient was admitted for treatment with coagulation factors and pain management. Dental procedures and local anaesthesia in patients with a severe coagulation disorder require specific preparatory measures, such as administration of coagulation factors. Collaboration and consultation with a patient's haematologist or haemophilia treatment centre are essential requirements for safe dental care.

Background: Desmopressin is frequently used perioperatively in persons with nonsevere hemophilia A. However, increase in factor (F)VIII:C after desmopressin use is interindividually highly variable. Tachyphylaxis has only been reported in test setting for persons with hemophilia A, with a remaining response of approximately 70% after a second dose compared with that after a first dose., Objectives: To study tachyphylaxis of FVIII:C response after multiple administration(s) of desmopressin in perioperative persons with nonsevere hemophilia A., Methods: We studied FVIII:C levels after desmopressin before (day 0 [D0]) and on days 1 (D1) and 2 (D2) after surgery in 26 patients of the DAVID and Little DAVID studies. We studied tachyphylaxis by comparing the responses at D1 and D2 with that at D0. We also assessed the reproducibility of the D0 response in comparison to an earlier performed desmopressin test., Results: The median absolute FVIII:C increase was 0.50 IU/mL (0.35-0.74; n  = 23) at D0, 0.21 IU/mL (0.14-0.28; n  = 17) at D1, and 0.23 IU/mL (0.16-0.30; n  = 11) at D2. The median percentage of FVIII increase after the second administration (D1) compared with the first (D0) was 42.9% (29.2%-52.5%; n  = 17) and that of the third (D2) compared with the first (D0) was 36.4% (23.7%-46.9%; n  = 11). The FVIII:C desmopressin response at D0 was comparable with the desmopressin test response in 74% of the patients., Conclusion: Tachyphylaxis in the surgical setting was considerably more pronounced than previously reported, with FVIII:C at D1 and D2 of 36% to 43% of the initial response. Our results may have important implications for monitoring repeated desmopressin treatment when used perioperatively., (© 2024 The Authors.)

Purpose: Pulmonary emboli (PE) contribute substantially to coronavirus disease 2019 (COVID-19) related mortality and morbidity. Immune cell-mediated hyperinflammation drives the procoagulant state in COVID-19 patients, resulting in immunothrombosis. To study the role of peripheral blood mononuclear cells (PBMC) in the procoagulant state of COVID-19 patients, we performed a functional bioassay and related outcomes to the occurrence of PE. Secondary aims were to relate this functional assay to plasma D-dimer levels, ventilation perfusion mismatch and TF expression on monocyte subsets., Methods: PBMC from an ICU biobank were obtained from 20 patients with a computed tomography angiograph (CTA) proven PE and compared to 15 COVID-19 controls without a proven PE. Functional procoagulant properties of PBMC were measured using a modified fibrin generation time (MC-FGT) assay. Tissue factor (TF) expression on monocyte subsets were measured by flow cytometry. Additional clinical data were obtained from patient records including end-tidal to arterial carbon dioxide gradient., Results: MC-FGT levels were highest in the samples taken closest to the PE detection, similar to the end-tidal to arterial carbon dioxide gradient (ETCO2 - PaCO2), a measurement to quantify ventilation-perfusion mismatch. In patients without proven PE, peak MC-FGT relates to an increase in end-tidal to arterial carbon dioxide gradient. We identified non-classical, CD16 positive monocytes as the subset with increased TF expression., Conclusion: We show that the procoagulant state of PBMC could aid in early detection of PE in COVID-19 ICU patients. Combined with end-tidal to ETCO2 - PaCO2 gradient, these tests could improve early detection of PE on the ICU., Competing Interests: Declaration of Competing Interest All the authors declare no conflicts of interest., (Copyright © 2023 The Author(s). Published by Elsevier Inc. All rights reserved.)

Background: Even though antithrombotic therapy has probably little or even negative effects on the well-being of people with cancer during their last year of life, deprescribing antithrombotic therapy at the end of life is rare in practice. It is often continued until death, possibly resulting in excess bleeding, an increased disease burden and higher healthcare costs., Methods: The SERENITY consortium comprises researchers and clinicians from eight European countries with specialties in different clinical fields, epidemiology and psychology. SERENITY will use a comprehensive approach combining a realist review, flash mob research, epidemiological studies, and qualitative interviews. The results of these studies will be used in a Delphi process to reach a consensus on the optimal design of the shared decision support tool. Next, the shared decision support tool will be tested in a randomised controlled trial. A targeted implementation and dissemination plan will be developed to enable the use of the SERENITY tool across Europe, as well as its incorporation in clinical guidelines and policies. The entire project is funded by Horizon Europe., Results: SERENITY will develop an information-driven shared decision support tool that will facilitate treatment decisions regarding the appropriate use of antithrombotic therapy in people with cancer at the end of life., Conclusions: We aim to develop an intervention that guides the appropriate use of antithrombotic therapy, prevents bleeding complications, and saves healthcare costs. Hopefully, usage of the tool leads to enhanced empowerment and improved quality of life and treatment satisfaction of people with advanced cancer and their care givers., Competing Interests: Declaration of competing interest AAH reports grants, speaking, or consulting fees from Bayer, MSD, LEO-Pharma and Bristol-Myers Squibb/Pfizer. MS reports grants and consulting fees from Bayer. JJB is a speaker for Abbott and Edwards Lifesciences. GJG is vice-chair of the Dutch Federation of Thrombosis services, and reports grants from Bayer, BMS, Pfizer and Daiichi-Sankyo, The Dutch Research Council, The Netherlands Organisation for Health Research and Development, and the Dutch Heart Foundation, all unrelated to this work and paid to his institution. MVH reports grants from Dutch Heart Foundation, ZonMw, Bayer Health Care, Pfizer-BMS, and Leo Pharma, all unrelated to this work. FAK reports grants or contracts from Bayer, BMS, BSCI, MSD, Leo Pharma, Actelion, Varm-X, The Netherlands Organisation for Health Research and Development, the Dutch Thrombosis Association, The Dutch Heart Foundation and the Horizon Europe Program, all unrelated to this work and paid to his institution. AM reports advisory board honoraria from Bayer, BMS Sanofi and Pfizer, speaker's bureau for BMS and Bayer and an educational grant from Bayer. DM reports contracts from BMS, Leo Pharma, Pfizer and Sanofi. MJHAK reports grants from Sobi, The Netherlands Organisation for Health Research and Development and the Dutch Thrombosis Association and speakers fee from Roche, Sobi and BMS, all unrelated to this work and paid to his institution. FEMM reports grants from Yorkshire Cancer Research, Marie Curie Cancer Care, the UK National Institute of Health and Care Research, and Hull Clinical Commissioning Group, all unrelated to this work. LV reports a grant from the Dutch Cancer Society, the Netherlands Organization for Health Service Research and Development, and the Dutch Research Council, all unrelated to this work. All others report no conflicts of interest related to this project., (Copyright © 2023 The Authors. Published by Elsevier Ltd.. All rights reserved.)

Currently, there is no consensus on the optimal management to prevent postpartum hemorrhage (PPH) in hemophilia carriers. We aimed to evaluate peripartum management strategies in relation to maternal and neonatal bleeding outcomes by performing an extensive database search up to August 2020. Seventeen case-reports/series and 11 cohort studies were identified of overall 'poor' quality describing 502 deliveries. The PPH incidence in the individual patient data was 63%; 44% for those women receiving prophylaxis to correct coagulation and 77% for those without (OR 0.23, CI 0.09-0.58) and in cohort data 20.3% (26.8% (11/41) vs. 19.4% (55/284) (OR: 1.53, 95% CI: 0.72-3.24), respectively. Peripartum management strategies mostly consisted of clotting factor concentrates, rarely of desmopressin or plasma. Tranexamic acid appears promising in preventing secondary PPH, but was not used consistently. Neonatal bleeding was described in 6 affected male neonates, mostly after instrumental delivery or emergency CS, but insufficient information was provided to reliably investigate neonatal outcome in relation to management. The high PPH risk seems apparent, at most mildly attenuated by prophylactic treatment. Prospective cohort studies are needed to determine the optimal perinatal management in hemophilia., (Copyright © 2021 The Authors. Published by Elsevier Ltd.. All rights reserved.)

Introduction: In the first wave, thrombotic complications were common in COVID-19 patients. It is unknown whether state-of-the-art treatment has resulted in less thrombotic complications in the second wave., Methods: We assessed the incidence of thrombotic complications and overall mortality in COVID-19 patients admitted to eight Dutch hospitals between September 1st and November 30th 2020. Follow-up ended at discharge, transfer to another hospital, when they died, or on November 30th 2020, whichever came first. Cumulative incidences were estimated, adjusted for competing risk of death. These were compared to those observed in 579 patients admitted in the first wave, between February 24th and April 26th 2020, by means of Cox regression techniques adjusted for age, sex and weight., Results: In total 947 patients with COVID-19 were included in this analysis, of whom 358 patients were admitted to the ICU; 144 patients died (15%). The adjusted cumulative incidence of all thrombotic complications after 10, 20 and 30 days was 12% (95% confidence interval (CI) 9.8-15%), 16% (13-19%) and 21% (17-25%), respectively. Patient characteristics between the first and second wave were comparable. The adjusted hazard ratio (HR) for overall mortality in the second wave versus the first wave was 0.53 (95%CI 0.41-0.70). The adjusted HR for any thrombotic complication in the second versus the first wave was 0.89 (95%CI 0.65-1.2)., Conclusions: Mortality was reduced by 47% in the second wave, but the thrombotic complication rate remained high, and comparable to the first wave. Careful attention to provision of adequate thromboprophylaxis is invariably warranted., (Copyright © 2021 The Author. Published by Elsevier Ltd.. All rights reserved.)

Background: In recent years, more awareness is raised about sex-specific dilemmas in inherited bleeding disorders. However, no large studies have been performed to assess differences in diagnosis, bleeding phenotype and management of men and women with bleeding disorders. Therefore, we investigated sex differences in a large cohort of well-defined patients with autosomal inherited bleeding disorders (von Willebrand disease (VWD), rare bleeding disorders (RBDs) and congenital platelet defects (CPDs))., Methods: We included patients from three nationwide cross-sectional studies on VWD, RBDs and CPDs in the Netherlands, respectively the WiN, RBiN and TiN study. In all studies a bleeding score (BS) was obtained, and patients filled in an extensive questionnaire on the management and burden of their disorder., Findings: We included 1092 patients (834 VWD; 196 RBD; 62 CPD), of whom 665 (60.9%) were women. Women were more often referred because of a bleeding diathesis than men (47.9% vs 36.6%, p  = 0.002). Age of first bleeding was similar between men and women, respectively 8.9 ± 13.6 (mean ±sd) years and 10.6 ± 11.3 years ( p  = 0.075). However, the diagnostic delay, which was defined as time from first bleeding to diagnosis, was longer in women (11.6 ± 16.4 years) than men (7.7 ± 16.6 years, p  = 0.002). Similar results were found when patients referred for bleeding were analyzed separately. Of women aging 12 years or older, 469 (77.1%) had received treatment because of sex-specific bleeding., Interpretation: Women with autosomal inherited bleeding disorders are more often referred for bleeding, have a longer diagnostic delay, and often require treatment because of sex-specific bleeding., Funding: The WiN study was supported (in part) by research funding from the Dutch Hemophilia Foundation (Stichting Haemophilia), Shire (Takeda), and CSL Behring (unrestricted grant)., Competing Interests: F. Atiq received the CSL Behring-professor Heimburger Award 2018 and a travel grant from Sobi. J.L. Saes and S.E.M. Schols received travel support from Bayer, Takeda and Sobi. F.W.G. Leebeek received research support from CSL Behring and Shire/Takeda for performing the Willebrand in the Netherlands (WiN) study, and from UniQure and Sobi for other studies. He is consultant for UniQure, Novo Nordisk, BioMarin and Shire/Takeda, of which the fees go to the institution, and has received a travel grant from Sobi. He is also a DSMB member for a study by Roche. H.C.J. Eikenboom received research support from CSL Behring and he has been a teacher on educational activities of Roche. K.P.M. van Galen received unrestricted research support from CSL Behring and Bayer and speakers fee from Takeda, CSL Behring and Bayer. J.G. van der Bom has been a teacher on educational activities of Bayer. M.H. Cnossen has received grants from governmental research institutes, such as the Dutch Research institute (NOW), ZonMW, Innovation fund, NWO-NWA and unrestricted investigator initiated research grants as well as educational and travel funding from various companies over the years (Pfizer, Baxter/Baxalta/Shire, Bayer Schering Pharma, CSL Behring, Sobi Biogen, novo Nordisk, Novartis and Nordic Pharma), and has served as a member on steering boards of Roche and Bayer. All grants, awards and fees go to the Erasmus MC as an institution. The institution of K. Fijnvandraat has received unrestricted research grants from CSL Behring, Sobi and NovoNordisk and her institution received consultancy fees from Grifols, Takeda, Novo Nordisk and Roche. K. Meijer received research support from Bayer, Sanquin and Pfizer; speaker fees from Bayer, Sanquin, Boehringer Ingelheim, BMS and Aspen; consulting fees from Uniqure, of which all fees go to the institution. B.A.P. Laros‐van Gorkom has received unrestricted educational grants from Baxter and CSL Behring. M.J.H.A. Kruip received grants from governmental research institutes, such as the Dutch Research institute (ZonMW/NWO), Dutch Thrombosis Foundation, Innovation fund, unrestricted grants from Bayer, Pfizer, Daiichi Sankyo, Sobi and Boehringer Ingelheim and speakers fee from Bayer. R.E.G. Schutgens received grants from Bayer, grants from Baxalta, Pfizer, NovoNordisk. The institution of M. Peters received an unrestricted research grant from Pfizer. F.J.M. van der Meer received grants from CSL Behring, Pfizer, Bayer, Novo Nordisk, Sobi, Roche, and OctaPharma. W.L. van Heerde reports speaker and consultant and travel fees from Takeda, Bayer, CSL Behring, and Sobi. He is also founder and co-owner of Enzyre. None of the other authors has a conflict of interest to declare., (© 2021 The Authors.)