key: cord-0943766-5icsq7d9
authors: Matar, Reem H.; Than, Christian A.; Nakanishi, Hayato; Daniel, Rohan Suresh; Smayra, Karen; Sim, Bernice L.; Beran, Azizullah; Danoun, Omar A.
title: Outcomes of patients with thromboembolic events following coronavirus disease 2019 AstraZeneca vaccination: a systematic review and meta-analysis
date: 2022-01-03
journal: Blood Coagul Fibrinolysis
DOI: 10.1097/mbc.0000000000001113
sha: 8bf0516bdcea1d3e39cb30b9c1271396315f2a55
doc_id: 943766
cord_uid: 5icsq7d9

AstraZeneca coronavirus disease 2019 (COVID-19) vaccinations have recently been implicated in thromboembolism formations. Our aim was to investigate the outcomes of patients with thromboembolic events following the AstraZeneca vaccine (ChAdOx1 nCoV-19, AZD1222). A literature search was performed from December 2019 to September 2021. Eligible studies must report participants older than 18 years vaccinated with AstraZeneca and outcomes of thromboembolic events. Pooled mean or proportion were analyzed using a random-effects model. A total of 45 unique studies (number of patients = 144, 64.6% women, mean age 21–68 years) were included. The most common presenting adverse events were headache (12.1%), intracerebral hemorrhage (7.5%), and hemiparesis (7%). The most common thromboembolic adverse events were cerebral venous sinus thrombosis (38.5%) and deep vein thrombosis/pulmonary embolism (21.1%). The most common radiologic finding were intracerebral hemorrhage and cerebral venous thrombosis. Laboratory findings included thrombocytopenia (75%) and hypofibrinogenemia (41%). On admission, 64 patients tested positive for PF4-Heparin ELISA assay (80%). Seventy-four patients were hospitalized with 22 being admitted to the ICU. A total of 78 patients recovered while 39 patients died. This meta-analysis presents evidence to suggest vaccine-induced immune thrombotic thrombocytopenia (VITT) following AstraZeneca vaccine. Clinical practice must, therefore, account for the possibility of VITT and subsequent embolic events in certain individuals’ postvaccination with adenovirus-based COVID-19 vaccines. Serum anti-PF4 suggests diagnostic value for VITT and could subsequently inform treatment choices in such instances.

The first case of coronavirus was identified in Wuhan, China in early December 2019 [1] . The virus, part of the novel enveloped RNA betacoronavirus family, has been named as the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and its associated disease coronaviruses disease 2019 (COVID-19) [2] . Since then, COVID-19 has been declared as a pandemic affecting over 224 180 411 globally with more than 4 621 205 deaths [3] . As such, efforts have been directed to combat and manage this disease.

Currently, four companies (AstraZeneca-Oxford (Cambridge, United Kingdom) Pfizer-Biontech (New York, NY, USA and Mainz, Germany), Moderna (Cambridge, Massachusetts, USA), and Johnson and Johnson (New Brunswick, New Jersey, USA)) have manufactured vaccines that have been authorized for use in the European Union. Whilst all are yet to reach approval status, emerging data from double-blinded, randomized, controlled clinical trials have persuaded the American Food and Drug Administration (FDA) to permit emergency use authorization for the Pfizer-Biontech, Moderna and Johnson and Johnson vaccines, whilst the European Medicine Agency has permitted use of the AstraZeneca vaccine, and full FDA approval for Pfizer-Biontech vaccine. Two of these vaccines are messenger RNA-based vaccines -BNT162b2 (Pfizer-BioNTech) and mRNA-1273 (Moderna) -which encode the spike protein antigen of SARS-CoV-2, encapsulated in lipid nanoparticles [4, 5] . The other two vaccines, ChAdOx1 nCoV-19 (Astra-Zeneca) and Ad26.COV2.S (Johnson and Johnson/Janssen), are recombinant adenoviruses that encode the spike glycoprotein of SARS-CoV-2 [6] .

To date, approximately 560 261 011 of vaccines have been administered in the European Union and 5.88 billion

The initial search yielded 567 potentially relevant articles from which 45 unique studies involving 144 patients met the eligibility criteria. The details of study selection process are depicted in Supplementary Item 2, http:// links.lww.com/BCF/A120. The baseline characteristics of the included studies are comprehensively described in Table 1 . The mean age ranged from 21 to 68 years of which 93 patients were women.

Results of the quality assessment of all included studies are shown in Supplementary Table 1, http://links.lww.com/BCF/A124. All the case series were judged to be of good quality. The patients appeared to represent the whole experience of the investigator and the exposure and outcome were adequately ascertained, and the length of follow-up was adequate. Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.

The clinical characteristics of the patients are shown in Table 2 . Among the overall population, some patients had at least one coexisting illness; frequently reported illnesses included pollen allergy (n ¼ 6), hypothyroidism (n ¼ 4) hypertension (n ¼ 8), asthma (n ¼ 3), diabetes (n ¼ 2), and neurologic disorders (n ¼ 2). One patient had comorbidity of Von Willebrand Disease, factor V Leiden thrombophilia, and anticardiolipin antibodies [14] . One patient had a relevant history of deep vein thrombosis (DVT) [15] . Sixteen patients were on preexisting medication prior to presentation. Some common medications included antihypertension agents [16, 17] and thyroid hormone replacement agents [16, 18] . The use of contraceptive methods was indicated in 15 patients. Specifically, seven patients were on the contraceptive pill [14, 15, 17] , three patients were on hormone replacement therapy, two on hormonal intrauterine device (IUD) [14] , and three patients on contraceptive vaginal ring [17] Table 3 shows the radiologic and laboratory findings on admission. The most common imaging performed were CT (n ¼ 114) and MRI (n ¼ 38). The most common radiologic findings were intracranial hemorrhage and cerebral venous sinus thrombosis. 

The pooled mean of time to onset of first adverse event following vaccination was 8.468 days (95% CI 7.486-9.451; I 2 ¼ 79.42%) ranging from 0 to 20 days. A total of 604 adverse events were reported, 223 of which were thromboembolic events. Out of 223 thromboembolic events, 70 were central venous sinus thrombosis (CVST), 67 were pulmonary embolism (PE)/DVT, and the rest were classified as other thromboembolic events. The pooled rate of CVST was 38.5% (95% CI 0.309-0.466, I 2 ¼ 6.54%). The rate of PE/DVT was 21.1% (95% CI 0.168-0.255, I 2 ¼ 0%). Seventy-two patients presented to the Emergency Room following an adverse event with a pooled rate of 73.1% (95% CI 0.617-0.820, I 2 ¼ 0%). Seventy-four patients were hospitalized, of which 22 were admitted to the ICU. The pooled rate of patients hospitalized was 80% (95% CI 0.708-0.868, I 2 ¼ 0%) and pooled rate of ICU admission was 44.1% (95% CI 0.310-0.582, I 2 ¼ 0%). The pooled mean of time to hospitalization after vaccination was 10.065 days (95% CI 8.275-11.856, I 2 ¼ 57.93%).

Patients received various treatment modalities over the course of their stay. Twenty-one patients received platelet transfusions with a pooled rate of 34.5% (95% CI 0.251-0.452, I 2 ¼ 0%). Nineteen patients were treated with a craniectomy with a pooled rate of 33.5% (95% CI 0.244-0.441, I 2 ¼ 0%). Ten patients received low-molecular-weight heparin (LMWH) with a pooled rate of 24.6% (95% CI 0.171-0.340, I 2 ¼ 0%), whereas 12 patients 

The cause of death of the patients was unknown. As shown in Table 4 and Fig. 1) , headache (12.1%; 95% CI 0.095-0.154, I 2 ¼ 0%), intracerebral hemorrhage (7.5%, 95% CI 0.056-0.102, I 2 ¼ 0%), hemiplegia (7%; 95% CI 0.051-0.096, I 2 ¼ 0%), fever (6.6%; 95% CI 0.047-0.091, I 2 ¼ 0%), congestive edema (5.3%; 95% CI 0.036-0.075, I 2 ¼ 0%), visual impairment (5.7%; 95% CI 0.040-0.081, I 2 ¼ 0%), and ocular manifestations (6.1%; 95% CI 0.043-0.086; I 2 ¼ 0%) were the most common reported adverse events following vaccination after first dose.

Six studies demonstrated thromboembolic events following Johnson and Johnson vaccination [19, 20] . Seventeen patients were included in total, of which 15 were girls. The mean age of patients was 40.6 years (range: 24-48 years). Eleven patients had preexisting comorbidities, seven of which had obesity, one had asthma, one had depression, one had hyperlipidemia, and one of which had hypothyroidism [20] . Seventeen of the patients presented and were admitted to the emergency room with thrombocytopenia. One of these patients had fibrinogenemia. Four patients were clinically diagnosed with vaccine-induced thrombocytopenia. There were 15 thromboembolic events of CVST. Moreover, out of the 17 patients who presented,15 patients out of 16 tested positive for antibody PF4-Heparin. Of the entire cohort, patients presented with 122 adverse events, of which 38 were thromboembolic events. Of the 38 thromboembolic events, 15 were CVST events and 10 were PE/DVT events, 6 were splanchnic events, and the rest were other thromboembolic events. Out of recorded reporting for 11 of these patients, 3 died and 8 recovered [20] . The baseline and clinical characteristics, radiological and laboratory findings, and outcomes of reported adverse events are reported alongside the AstraZeneca studies within Tables 1-4, respectively.

The aim of this systematic review and meta-analysis was to investigate outcomes of thromboembolic events in patients following AstraZeneca vaccine. Forty-five

AstraZeneca studies reporting on thromboembolism as an adverse event post vaccination were included, with six Johnson and Johnson studies. Within this meta-analysis, the following has been supported regarding AstraZeneca vaccine and thromboembolic events: under 60-year-olds have been the predominant age group reporting adverse events; the female sex appears to experience more adverse events than male sex; thrombocytopenia and hypofibrinogenemia appear as consistent findings in studies that report laboratory results; PF4 antibodies appear commonly present within patient serum whenever investigated. To the authors' knowledge, this is the most recent meta-analysis to describe thromboembolism adverse events following AstraZeneca vaccine. In turn, this study may assist clinical practice in determining adenovirus COVID-19 vaccine eligibility and management.

Recently, it has become evident that adenovirus-based vector vaccines may cause autoimmune thrombosis similar to heparin-induced thrombocytopenia (HIT) [21] . This phenomenon has been termed vaccine-induced immune thrombotic thrombocytopenia (VITT) because of its shared serological profile of high antibodies to platelet factor 4 PF4-polyanion complexes, as well as clinical presentations to that of HIT in which platelet disruption leads to thrombosis development [14, 17, 21, 22] . Clinical findings of thrombocytopenia and hypofibrinogenemia can, therefore, support suspicions of VITT, with the current meta-analysis finding thrombocytopenia prevalent in 75% of all patients who had data reported, and hypofibrinogenemia in 41%. Quantifying the estimated frequency of VITT from the included studies remains difficult, as the novel phenomenon was unknown to authors of earlier studies, with unclear guidelines for diagnosis. Now, in diagnosing VITT, guidance from the American Heart Association/ American Stroke Association Stroke Council suggests complete blood counts with a peripheral smear, coagulation studies with prothrombin time, partial thromboplastin time, fibrinogen, D-dimer, and PF4 antibody ELISA [15, 23] . Results of this current meta-analysis support these suggestions as included AstraZeneca studies conducting these analyses found ELISA anti-PF4 to be positive in 80% of patients tested, with corresponding abnormal patient serum samples and prolonged coagulation values (Table 3) .

Upon VITT, patients have been described to present with CVST or with other arterial or venous clots [21] . The pathophysiology for thrombus formation has been previously described in detail, with the conclusion that cellular positive feedback signaling results in a hypercoagulable state post adenoviral vaccination [24] . Autoantibodies are also generated despite no heparin exposure, leading to theories of an unidentified polyanion in the adenoviral vaccine itself or infected cells causing binding to PF4 [24] . Considering this mechanism, and that prior Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved. Table 4 Reported adverse events following vaccination and common treatment modalities thrombosis is not currently considered a risk factor for VITT, preemptive thrombophilia screening may not yield clinically useful information in identifying VITT susceptibility [24] . However, future studies are required to confirm this.

Due to ambiguity surrounding the exact mechanisms of VITT, there are no prominent risk factors for VITT with thromboembolism post vaccination other than female sex and age younger than 60 years [24] , reflecting the demographics of patients included within the current study. It is, therefore, vital that clinicians take appropriate caution when administering adenovirus vaccination to this patient population. Outside of VITT-specific risk factors, standard thromboembolism risk factors of thrombophilia, pregnancy, the postpartum timeframe, and hormonal contraceptives are thought to apply in a general sense to patients receiving adenovirus vaccination [25, 26] . However, limited reporting on these factors precluded investigations within this current analysis.

Out of embolic events, CVST was the most common with a pooled value of 38.5% in line with current reporting of embolism cases to EudraVigilance [27] . The high prevalence of CVST in turn largely explains the results of corresponding adverse event presentations. The main clinical syndromes seen with CVST are intracranial hypertension presenting as headache, focal deficits with hemiparesis and fluent aphasia, seizures, and venous hemorrhage [28, 29] , all of which were prevalent presenting adverse events within this meta-analysis. Consequently, presentation of these adverse events post vaccination should immediately guide clinical decisionmaking towards a diagnosis of VITT-related thromboembolism. The pooled onset of initial adverse event symptoms appearing approximately 10 days after vaccination is in line with current literature reporting a similar timeframe after receiving the first vaccine dose [23, 30] . This highlights a clear delay in adverse event presentation post vaccination that clinicians must be aware of, with necessity for appropriate preemptive management after the first vaccine dose as compared with the second. Unfortunately, scarcity of literature precludes any further comment on outcomes of thromboembolic events following second dose vaccination.

AstraZeneca has not been the only COVID-19 vaccine to present with thromboembolic events, as six studies from the USA reporting embolic events following the Johnson and Johnson vaccine were included as a subgrouping within the current meta-analysis. All patients reported similar clinical pictures to that of AstraZeneca vaccine patients, in that headaches were the most common presenting adverse event, followed by thromboembolic events that saw a high prevalence of CVST. Additionally, tested patients were 93.8% positive for ELISA anti-PF4, similar to AstraZeneca reports ( Table 3 ). Given that both AstraZeneca ChAdOx1 nCov-19 and Johnson and Johnson/Janssen Ad26.COV2.S are nonreplicating adenovirus vector-based DNA vaccines [19] , it is expected that the clinical course and laboratory results would share similarities. Slight differences have been shown in delayed clinical manifestations for Ad26.COV2.S, with lower D-dimer and activated partial thromboplastin time levels [31] . However, all other clinical characteristics remain comparable [31] . Subsequently, the findings of this meta-analysis further indicate a similarity between these two vaccines in the pathogenesis of VITT leading to thromboembolic events [20] .

As of now, recommended treatment for confirmed VITT according to guidelines includes the avoidance of heparin (both unfractionated heparin and LMWH) and platelet transfusions [21] . Instead, the UK's expert Hematology panel has advised anticoagulating with nonheparin-based therapies depending on the patient's drug profile and situation [24, 32] . These include direct oral anticoagulants (DOACs), such as dabigatran, apixaban, rivaroxaban, edoxaban, and fondaparinux, as well as parenteral direct thrombin inhibitors (e.g. bivalirudin and argatroban) [24] . Furthermore, urgent use of high-dose intravenous immunoglobulin at rate of 1 g/kg of body weight daily for 2 days has also been suggested [21, 32, 33] . In delays of initiating intravenous immunoglobin, steroid administration has been advised [32] , whilst in cases of declining fibrinogen levels below 1.5 g/l, fibrinogen concentrate, or cryoprecipitate should be considered [32] . These treatment modalities were largely seen within the current meta-analysis, reflecting current developing clinical practice (Table 4) .

Whilst this meta-analysis reports on arising thromboembolic events, it should be noted that these cases have thus far been a rarity in opposition to the current widespread trials and live administrations, which have not yet reported such events [9, 30, 34, 35] . At present, approximately 21 400 000 Johnson and Johnson vaccines have been administered within the USA, whilst 500 000 000 Astrazeneca vaccines have been administered within Europe. The sample size included within this meta-analysis is, therefore, miniscule in comparison to patients safely vaccinated with adenoviral-based vaccines. In addition, the risk of CVST associated with COVID-19 infection is considerably greater than that associated with vaccination [36] , which should dissuade any vaccine hesitancy.

Considering the rapid pace at which both COVID-19 vaccine trials and adverse event outcome reporting is occurring, the limitations of this current systematic review and meta-analysis must be acknowledged. The most pressing of these is the lack of high-quality data in the included studies. Due to the urgent timeline for data extraction and the complications surrounding VITT, many cases had incomplete documentation of the epidemiological history, laboratory values, particularly about anti-PF4 levels, and outcomes including cause of

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

Epidemiology, genetic recombination, and pathogenesis of coronaviruses

World Health Organization. Coronavirus disease (COVID-19) situation reports

An evidence based perspective on mRNA-SARS-CoV-2 vaccine development

A thermostable mRNA vaccine against COVID-19

Immunogenicity and safety of a recombinant adenovirus type-5-vectored COVID-19 vaccine in healthy adults aged 18 years or older: a randomised, double-blind, placebo-controlled, phase 2 trial

Adverse events in healthcare workers after the first dose of ChAdOx1 nCoV-19 or BNT162b2 mRNA COVID-19 vaccination: a single center experience

Oxford COVID Vaccine Trial Group. Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK

Analysis of thrombotic adverse reactions of COVID-19 AstraZeneca vaccine reported to EudraVigilance database

Methodological quality and synthesis of case series and case reports

Meta-analysis in clinical trials

Thrombotic thrombocytopenia after ChAdOx1 nCov-19 vaccination

Pathologic antibodies to platelet factor 4 after ChAdOx1 nCoV-19 vaccination

Thrombocytopenia with acute ischemic stroke and bleeding in a patient newly vaccinated with an adenoviral vectorbased COVID-19 vaccine

Thrombosis and thrombocytopenia after ChAdOx1 nCoV-19 vaccination

Deep vein thrombosis more than two weeks after vaccination against COVID-19

Thrombotic thrombocytopenia after Ad26.COV2.S vaccination

US case reports of cerebral venous sinus thrombosis with thrombocytopenia after Ad26.COV2.S vaccination

Razak F et al on behalf of the Drugs & Biologics Clinical Practice Guidelines Working Group and the Ontario COVID-19 Science Advisory Table. Vaccine-induced prothrombotic immune thrombocytopenia (VIPIT) following AstraZeneca COVID-19 vaccination. Science Briefs of the Ontario COVID-19 Science Advisory Table

Heparin-induced thrombocytopenia

American Heart Association/American Stroke Association Stroke Council Leadership. Diagnosis and management of cerebral venous sinus thrombosis with vaccine-induced immune thrombotic thrombocytopenia

Clinical characteristics and pharmacological management of COVID-19 vaccine-induced immune thrombotic thrombocytopenia with cerebral venous sinus thrombosis: a review

Firstdose ChAdOx1 and BNT162b2 COVID-19 vaccines and thrombocytopenic, thromboembolic and hemorrhagic events in Scotland

Diagnosis and management of cerebral venous thrombosis

AstraZeneca's COVID-19 vaccine: EMA finds possible link to very rare cases of unusual blood clots with low blood platelets

Cerebral venous thrombosis: a comprehensive review

Cerebral venous thrombosis

Oxford COVID Vaccine Trial Group. Safety and immunogenicity of ChAdOx1 nCoV-19 vaccine administered in a prime-boost regimen in young and old adults (COV002): a single-blind, randomised, controlled, phase 2/3 trial

Comparison of vaccine-induced thrombotic events between ChAdOx1 nCoV-19 and Ad26.COV.2.S vaccines

ChAdOx1 nCOV-19 vaccine-induced immune thrombotic thrombocytopenia and cerebral venous sinus thrombosis (CVST)

Diagnosis and management of vaccine-related thrombosis following AstraZeneca COVID-19 vaccination: guidance statement from the GTH

Oxford COVID Vaccine Trial Group. Safety and immunogenicity of the ChAdOx1 nCoV-19 vaccine against SARS-CoV-2: a preliminary report of a phase 1/2, single-blind, randomised controlled trial

The precautionary principle, the AstraZeneca COVID-19 vaccine and mixed messaging

Cerebral venous thrombosis and portal vein thrombosis: A retrospective cohort study of 537,913 COVID-19 cases

We would like to thank Larry J. Prokop MS, MAR for the literature search. Data availability statement: with publication, the data set used for this meta-analysis will be shared upon request from the study authors Ethical approval: this systematic review and meta-analysis does not require ethical approval

Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.mortalities. In addition, despite pooled information providing clinical benefit, there was a reliance on case series or case reports arising within the preceding months of this meta-analysis being performed. Furthermore, vaccinated patients who were asymptomatic or had mild adverse events, and who did not require hospitalization, were not accounted for because of publication bias. This prevented calculations on true incidence rates of thromboembolism from evolving cases or discrepancies in reporting. Lastly, because of the nature of the virus and the urgent need for more studies, this meta-analysis might have missed emerging studies recently published in the literature, particularly in languages other than English.Nevertheless, this meta-analysis presents novel evidence of VITT-related thromboembolism following AstraZeneca vaccine in certain individuals. Female individuals under 60 years of age seem as the primary group affected; however, male individuals have also been shown to be as susceptible. Further investigations are warranted into the mechanisms of potential thromboembolic formation following adenovirus vector-based DNA vaccine administration and causes for predisposition towards CVST specifically. Such elucidations may identify suitability for adenovirus vaccination and better guide clinical practice. At present, serum anti-PF4 suggests diagnostic value for VITT, and can inform treatment choices.

There are no conflicts of interest.