Booster vaccination with SARS-CoV-2 mRNA vaccines and myocarditis in adolescents and young adults: a Nordic cohort study.

BACKGROUND AND AIMS: The SARS-CoV-2 mRNA vaccines are associated with an increased risk of myocarditis. This association appears to be strongest in male adolescents and younger males and after the second dose. The aim was to evaluate the risk of myocarditis following SARS-CoV-2 mRNA booster vaccination in 12-to-39-year-olds. METHODS: A multinational cohort study was conducted using nationwide register data in Denmark, Finland, Norway, and Sweden and comprising all 8.9 million individuals residing in each of the four countries. Participants were followed for an inpatient diagnosis of myocarditis. In each of the four countries, Poisson regression was used to estimate adjusted incidence rate ratios (IRRs) of myocarditis comparing vaccination schedules, with associated 95% confidence intervals (CIs). Country-specific results were combined in meta-analyses. RESULTS: A total of 8.9 million residents were followed for 12 271 861 person-years and 1533 cases of myocarditis were identified. In 12-to-39-year-old males, the 28-day acute risk period following the third dose of BNT162b2 or mRNA-1273 was associated with an increased incidence rate of myocarditis compared to the post-acute risk period 28 days or more after the second dose [IRR 2.08 (95% CI 1.31-3.33) and 8.89 (2.26-35.03), respectively]. For females, the corresponding IRR was only estimable for BNT162b2, 3.99 (0.41-38.64). The corresponding absolute risks following the third dose of BNT162b2 and mRNA-1273 in males were 0.86 (95% CI 0.53-1.32) and 1.95 (0.53-4.99) myocarditis events within 28 days per 100 000 individuals vaccinated, respectively. In females, the corresponding absolute risks following the third dose of BNT162b2 were 0.15 (0.04-0.39) events per 100 000 individuals vaccinated. No deaths occurred within 30 days of vaccine-related cases. CONCLUSIONS: The results suggest that a booster dose is associated with increased myocarditis risk in adolescents and young adults. However, the absolute risk of myocarditis following booster vaccination is low.

Socioeconomic factors associated with poor medication adherence in patients with type 2 diabetes.

PURPOSE: This study aims to determine initiation and persistence for patients with type 2 diabetes receiving their first prescription of an antidiabetic agent and the associations with socioeconomic factors. METHODS: A cohort study including 8515 patients with type 2 diabetes who were prescribed their first antidiabetic medication between 2012 and 2019 in Uppsala, Sweden, was followed during 2 years. Medical records were linked to national registers on dispensed drugs and socioeconomic data. Adherence was assessed based on patients' medication claims within 30 days of prescription (initiation) and continued claims after 24 months (persistence). Multivariable logistic regression was used to determine the associations with the socioeconomic factors age, sex, living status, country of birth, education, occupation, and income. RESULTS: Within 30 days, 92.4% of the patients claimed their first prescription, and 64.0% were still being dispensed the initially prescribed medication after 24 months. Unemployed patients had lower initiation rates, and women had lower persistence rates. Factors associated with both low initiation and persistence were low income, young or old age, birth outside Europe, and being prescribed other diabetes drugs than metformin monotherapy. CONCLUSION: Socioeconomic factors have different impact on the initiation of a new medication and the persistence to treatment in type 2 diabetes. It is important to acknowledge these differences to develop appropriate interventions to improve medication nonadherence.

Hoarding of insulin and oral antidiabetics in the first phase of the COVID-19 pandemic: Regulatory actions taken to avoid shortage.

OBJECTIVE: To report on hoarding of prescribed medicines, with focus on insulins, in the early phase of the COVID-19 pandemic and on regulatory actions taken to avoid shortage. MATERIALS AND METHODS: The National Prescribed Drug Register which utilizes the Anatomic Therapeutic Chemical (ATC) Classification System and covers the total Swedish population was used. We calculated the number of packages of insulins (ATC code A10A), oral anti-diabetics (A10B), and all medicines across all ATC codes combined (A-S) dispensed per week in 2019 and 2020. Correspondingly, the number of packages of glucose test strips dispensed was calculated using the data source Concise held by the Swedish eHealth Agency. RESULTS: Prompt increases in numbers of dispensed packages were observed in March, peaking at week 11/2020. The absolute numbers of packages dispensed in week 11/2019 and week 11/2020 were: insulin, 49,694 and 95,767, an increase by +92.7%; oral antidiabetics, 55,478 and 82,684, +47.1%; glucose test strips, 18,119 and 23,476, +29:6%; and all medicines across all ATC codes combined, 1,988,456 and 2,659,421, +33.7%. Voluntary restriction of dispensing and a rapid change to applicable regulation were implemented within 2 weeks. A steep decline occurred, which became more pronounced after temporary regulation came in force from April 1, then leveling out during the following months. CONCLUSION: A signal of insulin hoarding was detected early in the COVID-19 pandemic. A temporary regulation, reducing dispensing to a maximum supply of 3 months was rapidly implemented. A shortage of vitally important prescribed medicines was avoided.

Identification of risk factors for adverse drug reactions in a pharmacovigilance database.

INTRODUCTION: In addition to identifying new safety signals, pharmacovigilance databases could be used to identify potential risk factors for adverse drug reactions (ADRs). OBJECTIVE: To evaluate whether data mining in a pharmacovigilance database can be used to identify known and possible novel risk factors for ADRs, for use in pharmacovigilance practice. METHOD: Exploratory data mining was performed within the Swedish national database of spontaneously reported ADRs. Bleeding associated with direct oral anticoagulants (DOACs)-rivaroxaban, apixaban, edoxaban, and dabigatran-was used as a test model. We compared demographics, drug treatment, and clinical features between cases with bleeding (N = 965) and controls who had experienced other serious ADRs to DOACs (N = 511). Statistical analysis was performed by unadjusted and age adjusted logistic regression models, and the random forest based machine-learning method Boruta. RESULTS: In the logistic regression, 13 factors were significantly more common among cases of bleeding compared with controls. Eleven were labelled or previously proposed risk factors. Cardiac arrhythmia (e.g., atrial fibrillation), hypertension, mental impairment disorders (e.g., dementia), renal and urinary tract procedures, gastrointestinal ulceration and perforation, and interacting drugs remained significant after adjustment for age. In the Boruta analysis, high age, arrhythmia, hypertension, cardiac failure, thromboembolism, and pharmacodynamically interacting drugs had a larger than random association with the outcome. High age, cardiac arrhythmia, hypertension, cardiac failure, and pharmacodynamically interacting drugs had odds ratios for bleeding above one, while thromboembolism had an odds ratio below one. CONCLUSIONS: We demonstrated that data mining within a pharmacovigilance database identifies known risk factors for DOAC bleeding, and potential risk factors such as dementia and atrial fibrillation. We propose that the method could be used in pharmacovigilance for identification of potential ADR risk factors that merit further evaluation.

Modeling SHH-driven medulloblastoma with patient iPS cell-derived neural stem cells.

Medulloblastoma is the most common malignant brain tumor in children. Here we describe a medulloblastoma model using Induced pluripotent stem (iPS) cell-derived human neuroepithelial stem (NES) cells generated from a Gorlin syndrome patient carrying a germline mutation in the sonic hedgehog (SHH) receptor PTCH1. We found that Gorlin NES cells formed tumors in mouse cerebellum mimicking human medulloblastoma. Retransplantation of tumor-isolated NES (tNES) cells resulted in accelerated tumor formation, cells with reduced growth factor dependency, enhanced neurosphere formation in vitro, and increased sensitivity to Vismodegib. Using our model, we identified LGALS1 to be a GLI target gene that is up-regulated in both Gorlin tNES cells and SHH-subgroup of medulloblastoma patients. Taken together, we demonstrate that NES cells derived from Gorlin patients can be used as a resource to model medulloblastoma initiation and progression and to identify putative targets.

Influence of statin-potency on the risk of kidney disease - A nationwide cohort study using laboratory data.

PURPOSE: To estimate the risk of kidney disease in high-potency statin users compared to those treated with low-potency statins without history of kidney disease at statin initiation, linking the Swedish national healthcare registers and laboratory data. METHODS: Incident users of statins, ≥40 years of age, with estimated Glomerular Filtration Rate (eGFR) >60 ml/min/1.73 m2 and no diagnosis of kidney disease at treatment initiation were identified between 2006 and 2007 and then followed for 2-years. The outcome was the incidence of kidney disease identified by the presence of the diagnostic code in the healthcare registers or eGFR <60 ml/min/1.73 m2 . We estimated hazard ratios (HRs) and 95% confidence intervals (CIs) with adjusted and propensity score (PS)-matched Cox proportional hazards models. RESULTS: A total of 27 385 patients were identified, 25.2% of which treated with a high-potency statin. During the follow-up, 68 (0.25%) patients were identified with a diagnosis of kidney disease from the registers. The number increased to 2498 (9.1%) when the criteria of eGFR <60 ml/min/1.73 m2 was added. The adjusted HR of kidney disease in high-potency statin users was 1.14 (95%CI 1.03-1.25) compared to low-potency users; the result was unchanged after the PS approach. CONCLUSIONS: Adding information from laboratory data to those from the national health registers, a slightly increased risk for kidney disease was found in high-potency statin users without pre-existing kidney disease at treatment initiation compared to those treated with low-potency statins.

A molecularly distinct subset of glioblastoma requires serum-containing media to establish sustainable bona fide glioblastoma stem cell cultures.

Glioblastoma (GBM) is the most frequent and deadly primary malignant brain tumor. Hallmarks are extensive intra-tumor and inter-tumor heterogeneity and highly invasive growth, which provide great challenges for treatment. Efficient therapy is lacking and the majority of patients survive less than 1 year from diagnosis. GBM progression and recurrence is caused by treatment-resistant glioblastoma stem cells (GSCs). GSC cultures are considered important models in target identification and drug screening studies. The current state-of-the-art method, to isolate and maintain GSC cultures that faithfully mimic the primary tumor, is to use serum-free (SF) media conditions developed for neural stem cells (NSCs). Here we have investigated the outcome of explanting 218 consecutively collected GBM patient samples under both SF and standard, serum-containing media conditions. The frequency of maintainable SF cultures (SFCs) was most successful, but for a subgroup of GBM specimens, a viable culture could only be established in serum-containing media, called exclusive serum culture (ESC). ESCs expressed nestin and SOX2, and displayed all functional characteristics of a GSC, that is, extended proliferation, sustained self-renewal and orthotopic tumor initiation. Once adapted to the in vitro milieu they were also sustainable in SF media. Molecular analyses showed that ESCs formed a discrete group that was most related to the mesenchymal GBM subtype. This distinct subgroup of GBM that would have evaded modeling in SF conditions only provide unique cell models of GBM inter-tumor heterogeneity.

FBW7 suppression leads to SOX9 stabilization and increased malignancy in medulloblastoma.

SOX9 is a master transcription factor that regulates development and stem cell programs. However, its potential oncogenic activity and regulatory mechanisms that control SOX9 protein stability are poorly understood. Here, we show that SOX9 is a substrate of FBW7, a tumor suppressor, and a SCF (SKP1/CUL1/F-box)-type ubiquitin ligase. FBW7 recognizes a conserved degron surrounding threonine 236 (T236) in SOX9 that is phosphorylated by GSK3 kinase and consequently degraded by SCFFBW7α Failure to degrade SOX9 promotes migration, metastasis, and treatment resistance in medulloblastoma, one of the most common childhood brain tumors. FBW7 is either mutated or downregulated in medulloblastoma, and in cases where FBW7 mRNA levels are low, SOX9 protein is significantly elevated and this phenotype is associated with metastasis at diagnosis and poor patient outcome. Transcriptional profiling of medulloblastoma cells expressing a degradation-resistant SOX9 mutant reveals activation of pro-metastatic genes and genes linked to cisplatin resistance. Finally, we show that pharmacological inhibition of PI3K/AKT/mTOR pathway activity destabilizes SOX9 in a GSK3/FBW7-dependent manner, rendering medulloblastoma cells sensitive to cytostatic treatment.

Batch-normalization of cerebellar and medulloblastoma gene expression datasets utilizing empirically defined negative control genes.

MOTIVATION: Medulloblastoma (MB) is a brain cancer predominantly arising in children. Roughly 70% of patients are cured today, but survivors often suffer from severe sequelae. MB has been extensively studied by molecular profiling, but often in small and scattered cohorts. To improve cure rates and reduce treatment side effects, accurate integration of such data to increase analytical power will be important, if not essential. RESULTS: We have integrated 23 transcription datasets, spanning 1350 MB and 291 normal brain samples. To remove batch effects, we combined the Removal of Unwanted Variation (RUV) method with a novel pipeline for determining empirical negative control genes and a panel of metrics to evaluate normalization performance. The documented approach enabled the removal of a majority of batch effects, producing a large-scale, integrative dataset of MB and cerebellar expression data. The proposed strategy will be broadly applicable for accurate integration of data and incorporation of normal reference samples for studies of various diseases. We hope that the integrated dataset will improve current research in the field of MB by allowing more large-scale gene expression analyses. AVAILABILITY AND IMPLEMENTATION: The RUV-normalized expression data is available through the Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) and can be accessed via the GSE series number GSE124814. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Mesenchymal transition and increased therapy resistance of glioblastoma cells is related to astrocyte reactivity.

Grade IV astrocytoma/glioblastoma multiforme (GBM) is essentially incurable, partly due to its heterogenous nature, demonstrated even within the glioma-initiating cell (GIC) population. Increased therapy resistance of GICs is coupled to transition into a mesenchymal (MES) cell state. The GBM MES molecular signature displays a pronounced inflammatory character and its expression vary within and between tumors. Herein, we investigate how MES transition of GBM cells relates to inflammatory responses of normal astroglia. In response to CNS insults astrocytes enter a reactive cell state and participate in directing neuroinflammation and subsequent healing processes. We found that the MES signature show strong resemblance to gene programs induced in reactive astrocytes. Likewise, astrocyte reactivity gene signatures were enriched in therapy-resistant MES-like GIC clones. Variable expression of astrocyte reactivity related genes also largely defined intratumoral GBM cell heterogeneity at the single-cell level and strongly correlated with our previously defined therapy-resistance signature (based on linked molecular and functional characterization of GIC clones). In line with this, therapy-resistant MES-like GIC secreted immunoregulatory and tissue repair related proteins characteristic of astrocyte reactivity. Moreover, sensitive GIC clones could be made reactive through long-term exposure to the proinflammatory cytokine interleukin 1 beta (IL1ß). IL1ß induced a slow MES transition, increased therapy resistance, and a shift in DNA methylation profile towards that of resistant clones, which confirmed a slow reprogramming process. In summary, GICs enter through MES transition a reactive-astrocyte-like cell state, connected to therapy resistance. Thus, from a biological point of view, MES GICs would preferably be called 'reactive GICs'. The ability of GBM cells to mimic astroglial reactivity contextualizes the immunomodulatory and microenvironment reshaping abilities of GBM cells that generate a tumor-promoting milieu. This insight will be important to guide the development of future sensitizing therapies targeting treatment-resistant relapse-driving cell populations as well as enhancing the efficiency of immunotherapies in GBM. © 2019 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

LGR5 promotes tumorigenicity and invasion of glioblastoma stem-like cells and is a potential therapeutic target for a subset of glioblastoma patients.

Glioblastoma (GBM) is the most common and lethal primary malignant brain tumor which lacks efficient treatment and predictive biomarkers. Expression of the epithelial stem cell marker Leucine-rich repeat-containing G-protein coupled receptor 5 (LGR5) has been described in GBM, but its functional role has not been conclusively elucidated. Here, we have investigated the role of LGR5 in a large repository of patient-derived GBM stem cell (GSC) cultures. The consequences of LGR5 overexpression or depletion have been analyzed using in vitro and in vivo methods, which showed that, among those with highest LGR5 expression (LGR5high ), there were two phenotypically distinct groups: one that was dependent on LGR5 for its malignant properties and another that was unaffected by changes in LGR5 expression. The LGR5-responding cultures could be identified by their significantly higher self-renewal capacity as measured by extreme limiting dilution assay (ELDA), and these LGR5high -ELDAhigh cultures were also significantly more malignant and invasive compared to the LGR5high -ELDAlow cultures. This showed that LGR5 expression alone would not be a strict marker of LGR5 responsiveness. In a search for additional biomarkers, we identified LPAR4, CCND2, and OLIG2 that were significantly upregulated in LGR5-responsive GSC cultures, and we found that OLIG2 together with LGR5 were predictive of GSC radiation and drug response. Overall, we show that LGR5 regulates the malignant phenotype in a subset of patient-derived GSC cultures, which supports its potential as a predictive GBM biomarker. Copyright © 2018 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Influence of baseline low-density lipoprotein cholesterol values on statin therapy persistence.

PURPOSE: Persistence to statins is low, in part due to lack of perception of cardiovascular (CV) risk. High values of low-density lipoprotein cholesterol (LDL-C) might increase the motivation for patients to be persistent. We investigated whether the baseline LDL-C value influences the discontinuation of statin treatment in patients with and without previous CV events. METHODS: A cohort study was performed using information from the Swedish national registers concerning dispensed drugs, hospital contacts, cause of death, and socioeconomic status, and linked with data from clinical laboratories. Incident statin users 20 years of age or older and starting treatment between 2006 and 2007 were identified and followed for 1 year. Baseline LDL-C level was defined as the last available laboratory test result during 6 months before the index statin dispensing. Cox regression was used to study discontinuation and estimate the effect on persistence of the baseline LDL-C value adjusting for sex, age, income, comorbidity, previous CV events, type of prescriber, and country of birth. Subgroup analyses stratifying by previous CV events and by diagnosis of diabetes among subjects without previous CV events were performed. RESULTS: A total of 29,389 patients were identified; 35.4% had a previous CV event. A high baseline LDL-C value was associated with a lower discontinuation rate (hazard ratio (HR) 0.81, 95% confidence interval (CI) 0.72-0.91) in patients without previous CV events. When stratifying further by diabetes diagnosis, the association was confirmed only in patients without diabetes. No association between LDL-C and persistence was found in patients with previous CV events. CONCLUSIONS: High levels of LDL-C were positively associated with statin persistence in newly treated diabetes patients without previous CV events.

Duration of dual antiplatelet treatment with clopidogrel and aspirin in patients with acute coronary syndrome.

AIMS: Dual antiplatelet treatment (DAPT) is a cornerstone in the treatment of acute coronary syndrome (ACS) but the optimal treatment duration is unclear. We aimed to evaluate the effect of DAPT duration with clopidogrel and aspirin on the recurrence of ischaemic events and bleeding in a large, unselected ACS population. METHODS AND RESULTS: We performed a prospective, observational cohort study of patients in Sweden (n = 56 440) admitted for ACS, with prescribed DAPT and hospitalized between January 2006 and July 2010. Patients were obtained from the SWEDEHEART register and data were merged with registers from the National Board of Health and Welfare. Depending on dispensed clopidogrel tablets, patients were divided into groups based on DAPT duration with clopidogrel and aspirin (3 months: 84-100 clopidogrel tablets (t); >3 months: >100 t; 6 months: 168-200 t; >6 months: >200 t). For the combined primary endpoint, defined as all-cause death, stroke, or re-infarction, only patients with an uneventful first 3-month period (no death, stroke, re-infarction, bleeding, stent thrombosis, or revascularization) were included. The incidence of the primary endpoint was 45 events per 1000 person-years in the >3 months DAPT group compared with 65 events per 1000 person-years in the 3 months DAPT group [adjusted HR 0.84, 95% CI (0.75; 0.95)]. Bleeding was more common in the >3 months treatment group (adjusted HR 1.56, 95% CI (1.18; 2.07), but the number of events was small. For >6 vs. 6 months DAPT, the adjusted HR for the combined endpoint was 0.75 with 95% CI (0.59; 0.95). CONCLUSION: In this contemporary, large real-life ACS population, DAPT for more than 3 months compared with a shorter duration was associated with a lower risk of death, stroke, or re-infarction. TRIAL REGISTRATION: Clinicaltrials.gov (NCT01623700).

Family history of cardiovascular disease and influence on statin therapy persistence.

PURPOSE: Persistence to statins is low, presumably due to lack of perception of risk. Having a close relative suffering from cardiovascular disease (CVD) might increase persistence to statins. We investigated whether family history of CVD influences the discontinuation of statin treatment. METHODS: A population-based cohort study was performed. Swedish registers on dispensed drugs, hospitalization and cause of death were linked to the Multi-generation Register. Incident statin users 20-72 years of age were identified between 2006 and 2007 and followed until 30 June 2009. Family history of CVD was defined as the presence of relatives with a previous cardiovascular event. Cox regression was used to study discontinuation and estimate the effect of the family history of CVD, adjusting for gender, age, education, income, healthcare provider, prevention's type, birth's country and residence's county. Stratified analysis by type and severity of cardiovascular event was performed. RESULTS: A total of 86,002 patients were enrolled; 61.5 % had a family history of CVD. Discontinuation of statin therapy was not associated with family history of CVD (HR: 0.98; 95 % CI:0.96-1.01), except for patients with a family history of death from myocardial infarction (MI) (HR: 0.95 95 % CI:0.92-0.98). Young age, foreign background, low income, and statin for primary prevention and for secondary prevention when prescribed by a general practitioner were associated with higher risk of statin discontinuation. CONCLUSIONS: Having relatives suffering from CVD did not consistently influence the persistence to statin treatment. A family history of death from MI had a slight significant positive effect on statin persistence, though not clinically relevant.

Non-steroidal anti-inflammatory drugs and venous thromboembolism in women.

BACKGROUND: Non-steroidal anti-inflammatory drugs (NSAIDs) might increase the risk of venous thromboembolism (VTE), and risks might differ by type of NSAID. Compared with men, women have a higher incidence of VTE at younger age, and they more often use NSAIDs. OBJECTIVES: To assess risks of VTE in young and middle-aged women in association with use of NSAIDs. PATIENTS/METHODS: In a nationwide case­control study (Thrombo Embolism Hormone Study) performed in Sweden 2003­2009, we included as cases 1433 women, 18 to 64 years of age with a first time VTE. Controls were 1402 randomly selected women, frequency matched by age. Information was obtained by telephone interviews and DNA analyses of blood samples. We calculated adjusted odds ratios (ORs) with 95% confidence intervals (CIs) adjusting for degree of immobilization, chronic disease, smoking, body mass index, use of hormonal contraception, hormone therapy or other NSAIDs. RESULTS: Use of NSAIDs was not associated with increased risks of VTE (OR = 0.98, 95% CI 0.80­1.19). The OR was 0.88 for propionic acid derivatives (95% CI 0.72­1.10), 1.18 for acetic acid derivatives (95% CI 0.82­1.70) and 1.76 for coxibs (95% CI 0.73­4.27). For users of acetic acid derivatives and coxibs, the ORs increased by cumulative dose. Carriership of the prothrombin gene mutation or factor V Leiden had only minor effects on the results. CONCLUSIONS: We found no increased risks of VTE in association with use of NSAIDs. Users of high cumulative doses of acetic acid derivatives and coxibs had the highest risks, suggesting a relationship with cyclooxygenase selectivity and dose.

Multi-country rapid adverse drug event assessment: the Asian Pharmacoepidemiology Network (AsPEN) antipsychotic and acute hyperglycaemia study.

PURPOSE: To undertake a multi-country study to investigate the risk of acute hyperglycaemia with antipsychotic use. METHODS: Using a distributed network model with a common minimal data set, we performed a prescription sequence symmetry analysis (PSSA) to investigate the risk of acute hyperglycaemia associated with antipsychotic initiation. Incident insulin prescriptions were used as a proxy indicator of acute hyperglycaemia. Participating countries and population datasets included Australia (300,000 persons), Japan I (300,000 persons), Japan II (200,000 persons), Korea (53 million persons) Taiwan (1 million persons), Sweden (9 million persons), USA-Public (87 million persons) and USA-Private (47 million persons). RESULTS: Olanzapine showed a trend towards increased risk in most databases, with a significant association observed in the USA-Public database (Adjusted sequence ratio (ASR) = 1.14; 95% Confidence Interval (CI) 1.10-1.17) and Sweden (ASR = 1.53; 95% CI 1.13-2.06). Null or negative associations were observed for haloperidol, quetiapine and risperidone. CONCLUSION: Acute hyperglycaemia appears to be associated with olanzapine use, however, this effect was only observed in two large databases. Despite different patterns of utilization of both antipsychotics and insulin, PSSA analysis results for individual antipsychotic medicines were qualitatively similar across most countries. PSSA, used in conjunction with existing methods, may provide a simple and timely method further supporting multi-national drug safety monitoring.

BERT based natural language processing for triage of adverse drug reaction reports shows close to human-level performance.

Post-marketing reports of suspected adverse drug reactions are important for establishing the safety profile of a medicinal product. However, a high influx of reports poses a challenge for regulatory authorities as a delay in identification of previously unknown adverse drug reactions can potentially be harmful to patients. In this study, we use natural language processing (NLP) to predict whether a report is of serious nature based solely on the free-text fields and adverse event terms in the report, potentially allowing reports mislabelled at time of reporting to be detected and prioritized for assessment. We consider four different NLP models at various levels of complexity, bootstrap their train-validation data split to eliminate random effects in the performance estimates and conduct prospective testing to avoid the risk of data leakage. Using a Swedish BERT based language model, continued language pre-training and final classification training, we achieve close to human-level performance in this task. Model architectures based on less complex technical foundation such as bag-of-words approaches and LSTM neural networks trained with random initiation of weights appear to perform less well, likely due to the lack of robustness that a base of general language training provides.

A temporary regulation to manage an impending shortage due to extraordinary prescribing patterns of chloroquines observed during early phase of COVID-19 epidemic.

Background: Chloroquine and hydroxychloroquine (C/HC) received considerable international media attention due to anticipated treatment effect in COVID-19. This led to increased prescriptions threatening to generate product shortages for patients prescribed within approved indications.We evaluated effects of a temporary regulation mandating pharmacies to only dispense C/HC prescribed by physicians with defined specialties. Methods: Data from Region Stockholm, which include 2.4 out of 10 million Sweden's population, were used. Weekly time trends of prescriptions and requisitions of C/HC by prescriber's workplace during January to April 2020 were followed. Results: Numbers of unique individuals with filled prescriptions of chloroquine increased tenfold and of hydroxychloroquine more than threefold from January to March. In the first week of April, filled prescriptions of C/HC dropped. In the later weeks of April, the number of filled prescriptions was back at similar levels as before the SARS-CoV-2 outbreak.During January and February, specialists in rheumatology accounted for 686 out of all 979 prescriptions dispensed (70.1%) of C/HC. In March, a large proportion of prescriptions dispensed were from specialists not usually prescribing C/HC, and rheumatology accounted for 628 out of all 1,639 prescriptions (38.3%). In April, specialists in rheumatology accounted for 386 out of all 641 prescriptions dispensed (60.0%). Conclusion: After an observed increase in prescriptions of C/HC, a temporary regulation was introduced on 2nd April 2020 to reduce prescriptions from specialists not usually prescribing C/HC to avoid shortages for patients within approved indications. Subsequently, dispensed prescriptions decreased from April and remained at pre-COVID-19 levels thereafter.

Validation of myocarditis diagnoses in the Swedish patient register for analyses of potential adverse reactions to COVID-19 vaccines.

Background: Coronavirus disease 2019 (COVID-19) mRNA vaccines are associated with an increased risk of myocarditis using hospital discharge diagnoses as an outcome. The validity of these register-based diagnoses is uncertain. Methods: Patient records for subjects < 40 years of age and a diagnosis of myocarditis in the Swedish National Patient Register were manually reviewed. Brighton Collaboration diagnosis criteria for myocarditis were applied based on patient history, clinical examination, laboratory data, electrocardiograms, echocardiography, magnetic resonance imaging and myocardial biopsy. Poisson regression was used to estimate incidence rate ratios, comparing the register-based outcome variable to validated outcomes. Interrater reliability was assessed by a blinded re-evaluation. Results: Overall, 95.6% (327/342) of cases registered as myocarditis were confirmed (definite, probable or possible myocarditis according to Brighton Collaboration diagnosis criteria, positive predictive value 0.96 [95% CI 0.93-0.98]). Of the 4.4% (15/342) cases reclassified as no myocarditis or as insufficient information, two cases had been exposed to the COVID-19 vaccine no more than 28 days before the myocarditis diagnosis, two cases were exposed >28 days before admission and 11 cases were unexposed to the vaccine. The reclassification had only minor impact on incidence rate ratios for myocarditis following COVID-19 vaccination. In total, 51 cases were sampled for a blinded re-evaluation. Of the 30 randomly sampled cases initially classified as either definite or probably myocarditis, none were re-classified after re-evaluation. Of the in all 15 cases initially classified as no myocarditis or insufficient information, 7 were after re-evaluation re-classified as probable or possible myocarditis. This re-classification was mostly due to substantial variability in electrocardiogram interpretation. Conclusion: This validation of register-based diagnoses of myocarditis by manual patient record review confirmed the register diagnosis in 96% of cases and had high interrater reliability. Reclassification had only a minor impact on the incidence rate ratios for myocarditis following COVID-19 vaccination.