Article Text
Abstract
Objectives Current guidelines do not recommend direct oral anticoagulants (DOACs) to treat cerebral venous thrombosis (CVT) despite their benefits over standard therapy. We performed a systematic review to summarise the published experience of DOAC therapy in CVT.
Data sources MEDLINE, Embase and COCHRANE databases up to 18 November 2020.
Eligibility criteria All published articles of patients with CVT treated with DOAC were included. Studies without follow-up information were excluded.
Data extraction and synthesis Two independent reviewers screened articles and extracted data. A risk of bias analysis was performed.
Primary and secondary outcome measures Safety data included mortality, intracranial haemorrhage (ICH) or other adverse events. Efficacy data included recurrent CVT, recanalisation rates and disability by modified Rankin Scales (mRS).
Results 33 studies met inclusion criteria. One randomised controlled trial, 5 observational cohorts and 27 case series or studies reported 279 patients treated with DOAC for CVT: 41% dabigatran, 47% rivaroxaban, 10% apixaban and 2% edoxaban, in addition to 315 patients treated with standard therapy. The observational cohorts showed a similar risk of death in DOAC and standard therapy arms (RR 2.12, 95% CI 0.29 to 15.59). New ICH was reported in 2 (0.7%) DOAC-treated patients and recurrent CVT occurred in 4 (1.5%). A favourable mRS between 0 and 2 was reported in 94% of DOAC-treated patients, more likely than standard therapy in observational cohorts (RR 1.13, 95% CI 1.02 to 1.25).
Conclusion The evidence for DOAC use in CVT is limited although suggests sufficient safety and efficacy despite variability in timing and dose of treatment. This systematic review highlights that further rigorous trials are needed to validate these findings and to determine optimal treatment regimens.
- stroke
- stroke medicine
- anticoagulation
- epidemiology
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
We performed an all-encompassing review of patients treated with direct oral anticoagulant (DOAC) for cerebral venous thrombosis (CVT).
Given the heterogeneity of the literature, a risk of bias analysis was performed.
We compared DOAC and standard therapy in one randomised controlled trial and five observational cohorts.
Meta-analysis comparing different DOACs was not possible and is a limitation of this study.
Introduction
Cerebral venous thrombosis (CVT) requires rapid treatment to prevent neurological disability or death due to venous infarct and haemorrhage. The estimated incidence is 1 per 100 000 per year with a mean age of onset 39 years.1 Although the mortality rate has reduced to 5%–15% due to advances in detection and treatment, morbidity rates can reach as high as 20%–30%.2 A Cochrane review in 2011 showed anticoagulation to be safe in CVT and was associated with a reduction in death prompting international guidelines to recommend acute treatment of CVT with either unfractionated heparin (UFH) or low molecular weight heparin (LMWH).3–6 Longer term anticoagulation is required since recurrent venous thromboembolism (VTE) is highest within the first year of CVT.7 Thus, at least 3 months of ongoing anticoagulation in low-risk patients and indefinitely for unprovoked, high-risk patients, or those with malignancy, is recommended.6 8 The transition from acute treatment of CVT with LMWH or UFH to an oral anticoagulant, such as warfarin, is standard practice despite no randomised controlled trial (RCT) comparing warfarin with UFH or LMWH.
Direct oral anticoagulants (DOACs) were introduced to treat symptomatic VTE over the past 10 years and have advantages over warfarin: more predictable pharmacokinetics, no international normalised ratio (INR) monitoring requirement or daily dose adjustments, while demonstrating similar efficacy in treatment of acute VTE with lower rates of intracranial haemorrhage (ICH).9 Guideline recommendations, however, do not support DOAC treatment for CVT given the paucity of evidence.6 Recent larger studies on DOAC therapy for VTE in atypical locations included CVT, thus assessment of the appropriateness of these anticoagulants for the treatment of CVT is warranted.10–12
The objective of this study was to review all available evidence to assess data on safety and efficacy of DOACs in the treatment of CVT.
Methods
Search strategy and selection criteria
The protocol for this systematic review was registered (PROSPERO ID: CRD42017078398)13 and published.14 We followed Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols,15 Preferred Reporting Items for Systematic Reviews and Meta-Analyses16 and Synthesis without meta-analysis (SWiM)17 guidelines where applicable. The search strategy was iteratively developed with assistance of a research librarian (RS) and is available in the supplement (online supplemental appendix 1. We searched Ovid MEDLINE, Embase and the Cochrane Central Register of Controlled Trials for original reports of patients with a diagnosis of CVT treated with a DOAC up to 18 November 2020. We included all available peer-reviewed studies including RCTs, prospective or retrospective observational cohorts, case series and case studies. Studies without follow-up data were excluded. Two authors (GB and JG) independently reviewed titles and abstracts for inclusion.
Supplemental material
Data items
Study type and number of patients were collected. Patient data included age, sex and medical history; CVT information included location of venous thrombosis and ICH; and DOAC data included type, dosage, timing of initiation after immediate therapy and duration of treatment. Safety outcomes included mortality, occurrence of intracranial and extracranial bleeding as defined by authors and any other reported adverse events. Efficacy outcomes included recurrent CVT, recanalisation rates and disability measured by the modified Rankin Scale (mRS). The mRS is a six-point scale ranging from 0 (no symptoms) to 6 (death), with a score of 2 indicating slight disability but able to look after own affairs without assistance.18 When applicable, authors were contacted for further data.
Risk of bias analysis
We used the Cochrane Risk of Bias Tool for randomised trials;19 the Newcastle Ottawa Scale for observational cohorts;20 and Joanna Briggs Institute (JBI) Critical Appraisal Checklist for case studies and case series.21 The Grading of Recommendations, Assessment, Development and Evaluations (GRADE) framework was used to assess the certainty of absolute treatment effects.22
Statistical analysis
Data were reported as counts and proportions for dichotomous data, medians and ranges for non-normally distributed continuous data, or means with SD for normally distributed continuous data. We reported risk ratios (RRs) with 95% CIs and study heterogeneity (I2) wherever possible. Case series and case report outcomes are presented as pooled descriptive statistics for each DOAC. Statistics were performed using STATA/IC V.15.1 and RevMan V.5.4.1.
Patient and public involvement
This systematic review had no individual patient involvement.
Results
Search results
Of 1843 titles, 33 studies met inclusion criteria (figure 1), representing 279 patients with CVT treated with a DOAC listed in table 1. We identified one RCT consisting of 60 patients treated with dabigatran and 60 patients treated with warfarin;23 five observational cohorts of 101 patients treated with rivaroxaban (n=80), dabigatran (n=11) and apixaban (n=10) compared with warfarin (n=301) or LMWH (n=14);24–28 six case series of patients treated with rivaroxaban (n=44), dabigatran (n=36) and apixaban (n=13);29–34 and 21 case studies of rivaroxaban (n=8), dabigatran (n=8), apixaban (n=4) and edoxaban (n=5).35–55 The clinical characteristics and outcomes of the patients are listed in table 2.
Dabigatran
A total of 115 patients (41.2%) were treated with dabigatran. In a multicentre, open-label, blinded end-point RCT by Ferro et al, ‘A Clinical Trial Comparing Efficacy and Safety of Dabigatran Etexilate With Warfarin in Patients With Cerebral Venous and Dural Sinus Thrombosis’ (RE-SPECT CVT, NCT02913326)23 patients were initially treated with LMWH or UFH for 5–15 days, followed by dabigatran 150 mg twice daily for 24 weeks. No patient died in the study. No new ICH occurred in the dabigatran group, while two occurred in the warfarin group. There were seven patients (11.7%) who discontinued dabigatran due to adverse events: one for worsening CVT-related baseline ICH, one intestinal haematoma and five non-bleeding adverse events. None of the four (6.7%) patients who discontinued warfarin did so due to adverse events. Follow-up data on 55 dabigatran-treated patients showed no radiographic CVT improvement in 40%, compared with 33% treated with warfarin (RR 1.22, 95% CI 0.74 to 2.03, p=0.44). At 24 weeks, a favourable mRS of 0–2 was reported in 58 of 59 (98.3%) in the dabigatran group and 56 of 58 (96.6%) in the warfarin group (p=0.62).
Descriptive studies of dabigatran reported an additional 44 patients. A case series by Mendonça et al33 provided patient-level data on request for 18 patients treated initially with UFH for a median 13 days followed by dabigatran for a median 6 months, 150 mg twice daily in 16 patients (89%) and 110 mg twice daily in two patients (11%). No deaths or ICH were reported, though one patient (6%) had a major intestinal bleed and one (6%) had minor intestinal bleed. At 6 months, mRS of 0 or 1 was reported in 15 patients (83%) and one (6%) had mRS of 3 (moderate disability, dependent on others but can walk). Rusin et al31 reported pooled data on 18 patients with dabigatran, 150 mg twice daily in 16 and 110 mg twice daily in 2, as well as rivaroxaban 20 mg daily in 10 and apixaban 5 mg twice daily in eight patients treated for a median of 8.5 months. During the 30-month follow-up, no death or ICH was reported, but three (8.3%) had major bleeding. Recurrent CVT occurred in two (5.6%) at 5 and 20 months after DOAC completion. Complete recanalisation occurred in 10 on dabigatran (55.6%), 6 on rivaroxaban (60.0%) and 6 on apixaban (50.0%). At 6–12 months after CVT, an excellent mRS of 0 or 1 was reported in 24 patients (66.7%), independent mRS of 2 in 10 (27.8%) and two (5.6%) had significant disability. Case studies of dabigatran reported one new ICH due to development of a dural arteriovenous fistula (DAVF) despite a reportedly complete recanalisation of their CVT37 and one myocardial infarction in the context of double thrombophilia from both plasminogen activator inhibitor-1 (PAI-1) 4G/4G homozygous genotype and protein C and S deficiency and required transition to warfarin.39 Otherwise, no patient had reported mortality, and all eight case studies reported an mRS of 0 or 1 after treatment.37–39 52–55
Rivaroxaban
A total of 132 patients (47.3%) were treated with rivaroxaban. Five observational cohorts pooled 101 DOAC-treated patients, 80 (79%) on rivaroxaban, 11 (11%) on dabigatran 150 mg twice daily and 10 (10%) on apixaban, compared with 315 on standard therapy with 301 (96%) warfarin and 14 (4%) LMWH.24–28 Patients were treated with DOAC for an average 8.1 months and with standard therapy for 9.8 months. Deaths were reported in four patients treated with a DOAC compared with six on standard therapy (RR 2.12, 95% CI 0.29 to 15.59, p=0.46, I2=49%) (figure 2). Hsu et al24 reported two deaths after DOAC therapy (25%): one in hospital from respiratory failure postaspiration in a patient treated with apixaban, and another due to metastatic lung cancer 1 year after CVT. Wasay et al27 reported two deaths in their DOAC group (4%): one prior to discharge and one prior to 6-month follow-up, and four deaths in their warfarin group (6%): three prior to discharge and one prior to 6-month follow-up. The causes of death were not reported. Herweh et al28 reported two deaths in their cohort (2%), and on request for patient-level data, none were treated with a DOAC. No significant difference between DOAC or standard therapy was reported for ICH (1% vs 2.5%, RR 0.72, 95% CI 0.18 to 2.85, p=0.64, I2=0%), recurrent CVT (5.7% vs 11.7%, RR 0.45, 95% CI 0.05 to 4.40, p=0.49, I2=54%) or incomplete recanalisation (35.8% vs 26.5%, RR 0.84, 95% CI 0.58 to 1.21, p=0.35, I2=0%) available in the supplement (online supplemental appendix 2). A favourable functional outcome of mRS 0–2 was reported in 61 of 69 (88.4%) DOAC-treated patients compared with 126 of 156 (80.7%) on standard therapy (RR 1.13, 95% CI 1.02 to 1.25, p=0.02, I2=0%) (figure 3).
Supplemental material
Descriptive studies of rivaroxaban reported an additional 52 patients. A case series by Shankar Iyer et al30 treated 20 stable patients with rivaroxaban acutely at 15 mg twice daily for 3 weeks followed by 20 mg daily. At 6-month follow-up, no patient died or discontinued rivaroxaban. There was no ICH or adverse effects reported. There was recanalisation in all patients and 19 (95%) reported mRS of 0 or 1, with mRS of 2 in only one (5%). Other case series and studies of rivaroxaban reported no mortality or ICH, and all had mRS 0 or 1 at follow-up.32 34–36 47–51 The dosing of rivaroxaban was variable: most received 20 mg daily after initial standard therapy,32 one with antiphospholipid syndrome received 15 mg daily after suffering a stroke with haemorrhagic transformation 3 months after starting warfarin for CVT,35 two received 10 mg daily in the context of Crohn’s disease49 and pegylated asparaginase for acute lymphoblastic leukaemia,48 and one was treated with 5 mg daily, in conjunction with plasma exchange (PLEX), for concurrent anti-N-methyl-D-aspartate (NMDA) receptor encephalitis.50 One patient was initially treated with rivaroxaban 15 mg twice daily and was then switched to dabigatran due to low anti-Xa levels in the context of concurrent phenytoin use for seizures secondary to CVT.52
Apixaban
Apixaban has been reported in 27 patients (9.7%).29 40 41 In the series reported by Covut et al29, five patients were treated with apixaban and four patients with rivaroxaban after a median 3 days of UFH and continued for a median of 12 months. No patient died or had new ICH during the follow-up, nor switched off their DOAC. One patient was switched onto apixaban due to gastrointestinal bleeding on warfarin and another was switched onto rivaroxaban 30 days after starting warfarin due to INR fluctuations. No recanalisation was reported in three patients (60%) on apixaban and one patient (25%) on rivaroxaban. At 6-month follow-up, mRS was 0 or 1 in eight patients (89%) and one patient had persistent mRS of 4 (unable to walk unassisted). The other case studies of apixaban indicate that all four patients had mRS of 0–1 after treatment, with no mortality or new ICH. Apixaban dosing was 5 mg twice daily for all patients, though one received 10 mg twice daily initially for 7 days in the context of T cell acute lymphoblastic leukaemia treated with pegylated asparaginase.40
Edoxaban
Edoxaban was reported in case studies of five patients (1.8%).37–41 No death, ICH, recurrent CVT or incomplete recanalisation was reported, and all patients had a good functional outcome. Two of the reported patients developed CVT in the context of COVID-19 infection and recovered without neurological sequelae.45 46
Risk of bias
The risks of bias analyses are available in the supplement (online supplemental appendix 3). In RE-SPECT CVT, patients and treating teams were aware of treatment allocation.23 No observational cohort controlled for confounders. Treatment initiation time was not reported in two observational cohorts, and follow-up duration was not standardised.24 26 The case series and case studies are moderately biased based on JBI Critical Appraisal, given lack of reporting completeness. Based on the currently available studies, the GRADE certainty is low for the absolute treatment effect.
Supplemental material
Discussion
We found that since the approval of DOAC for treatment of VTE, 279 patients treated with DOAC for CVT have been published with follow-up data. Of these patients, 42% are reported in case studies or case series, 36% in five observational cohorts and 22% in one RCT. There were 200 patients (72%) published in 2019 and 2020, suggesting that practitioner comfort for DOAC use in CVT is improving despite a lack of guideline recommendations.6 A recent survey of Canadian neurologists and haematologists suggests interest in the utilisation of DOAC for treatment of CVT, and the increasing reports support this trend.56
Outcomes of DOAC compared with standard therapy
Currently, warfarin is supported by guidelines despite no RCT evidence of superiority or non-inferiority to LMWH or UFH. The benefits of the DOAC over warfarin include reduced dose adjustments due to drug and food interactions, no need for INR monitoring to ensure therapeutic range and, in the case of dabigatran, the availability of a reversal agent. Furthermore, even when closely monitored in a clinical trial setting, patients on warfarin for CVT were in the therapeutic INR range only 66% of the time,23 suggesting better anticoagulation may be achieved with DOAC. Overall safety of DOAC was reassuring, with recurrent CVT, new ICH and death only reported in observational cohorts at rates similar to standard therapy and within the expected range of treated CVT.2 Furthermore, of the DOAC-treated patients who died, two of four deaths occurred after discharge, including one related to underlying metastatic cancer that would not suggest DOAC-related mortality.24 Efficacy was also promising with 93% of DOAC-treated patients attaining a favourable outcome of mRS from 0 to 2 compared with 85% of those on standard therapy. Compared with standard therapy in the observational cohorts, this value was higher for DOAC-treated patients. However, utilisation of DOAC in less severe CVT cannot be ruled out as a confounding factor since the observational cohorts did not have comparable standard treatment groups.
A meta-analysis published by Lee et al57 showed similar results to our review with no difference between DOAC or warfarin for recanalisation rates or major bleeding; however, their review analysed an ‘excellent’ mRS outcome of 0–1 and found no difference, while our study analysed a ‘favourable’ mRS of 0–2 and found a difference in the observational cohorts. The dichotomy of a favourable mRS has been debated, with mRS greater than two shown to be related to 1-year mortality, as well as being an independence cut-off for entry to certain endovascular trials.58–60 The apparent discrepancy may also relate to two of their analysed observational cohort studies (Geisbüsch et al61 and Herweh et al28) potentially including patients from the same institution during overlapping time periods (January 2012–December 2013 and January 1998–September 2014, respectively). To clarify, we were able to contact the authors from these studies and obtain patient-level data, which led to the exclusion of Geisbüsch et al61 due to duplicate patient data. Furthermore, we have updated the search to include an additional two cohorts published in 2020.
An ongoing RCT out of University of British Columbia, the ‘Study of Rivaroxaban for CeREbral Venous Thrombosis’ (SECRET, NCT03178864), is currently recruiting an estimated 50 participants comparing rivaroxaban with standard anticoagulation of LMWH, UFH or warfarin, expected to be completed December 2021.62 Another RCT, ‘Rivaroxaban vs Warfarin in CVT Treatment’ (RWCVT, NCT04569279) out of Damascus University has completed enrollment of 71 patients though not yet published results.63 Results of these studies will be useful for future guideline recommendations for DOAC use in CVT compared with standard therapy.6
Comparison between different DOAC
Our search yielded no randomised trials comparing different DOAC against each other, thus no formal meta-analysis comparing different DOAC was possible. Dabigatran was compared against warfarin in the only published RCT specifically looking at CVT to date; however, the most commonly reported DOAC was rivaroxaban, possibly suggesting physician comfort with this medication. Results from RWCVT and SECRET will help validate safety and efficacy of rivaroxaban and allow more definitive comparison with dabigatran from RE-SPECT CVT.62
The timing of DOAC initiation after acute treatment with LMWH or UFH ranged from 5 to 15 days for the RCT and from 3 to 12 days for the observational cohorts. The descriptive studies had more variability in DOAC initiation, ranging from acutely after CVT diagnosis, to as far as 3 months, making comparisons challenging. The dosage of DOAC was also inconsistent, with dabigatran dose ranging from 75 mg to 150 mg twice daily in the cohort by Wasay et al27 and rivaroxaban dosing between 5 mg daily and 20 mg daily depending on the study. Both ongoing RCTs use rivaroxaban after initial acute therapy with LMWH or UFH, for SECRET 20 mg daily within 14 days of CVT diagnosis, and for RWCVT 20 mg or 15 mg, depending on creatinine clearance, after a non-specified duration of acute therapy. These and future trials should help standardise how long initial therapy with LMWH or UFH is needed, if at all, prior to using DOAC, as well as if initial dosage adjustments are needed.
There were rare adverse events with each DOAC therapy. For dabigatran, no deaths were reported, and of the patients who experienced bleeding, none were given the reversal agent. However, in RE-SPECT CVT, dabigatran was stopped in two patients due to intestinal haematoma and worsening of the haemorrhagic component of their baseline intracranial lesion.23 Bleeding events on rivaroxaban were only reported in the series by Rusin et al31 in three patients (8.3%), two on 20 mg daily rivaroxaban and one on 110 mg twice daily dabigatran, who had heavy menstrual bleedings in two and upper gastrointestinal bleeding in one. Other rare adverse events include the in-hospital death of a patient treated with apixaban who had an aspiration event and respiratory failure,24 myocardial infarction while on dabigatran39 and DAVF formation 3 months after CVT despite complete recanalisation with dabigatran.37 A post hoc analysis of the RE-SPECT CVT showed no DAVF formation at 6 months.64 Two case studies of edoxaban treated patients with CVT in the context of COVID-19.45 46 Thrombotic complications of COVID-19 has been reported, but the safety and efficacy of DOAC in COVID-19 related thrombosis specifically has yet to be confirmed.65 66
The efficacy of each DOAC was good for treatment of CVT. Recurrent CVT was only reported in four patients overall (1.5%), two patients from the cohort Powell et al25 (11%) and two in the case series Rusin et al31 (5.6%) after discontinuation of DOAC. An international long-term cohort found the rate of recurrent CVT is as high as 4.4% at median 40 months; therefore, long-term follow-up of DOAC-treated CVT is needed to determine the ideal treatment duration.67 Recanalisation rates varied between DOAC treatment at similar rates reported in randomised trials of LMWH and UFH to treat CVT3–5 without clear reduction of a favourable functional outcome, as previously demonstrated.28 However, the prognostic value of recanalisation has been investigated by a meta-analysis of standard therapy, which showed recanalisation occurred in up to 85% of patients and was associated with mRS 0 or 1 (OR 3.3, 95% CI 1.7 to 6.3, p=0.001).68 Further high-quality studies will be required to determine if recanalisation rates differ between DOACs, as well as if they are related to functional outcome.
Limitations
The results of this systematic review should be interpreted with caution. The majority of patients were reported in retrospective observational cohorts or case studies prone to selection bias, confounding and lack of standardisation in timing of therapy initiation and follow-up duration. Therefore, pooling and inferential statistical analysis was not prudent due to the clinical and methodological heterogeneity and conclusions as to how DOAC therapies perform against each other could not be made. The risk of bias analysis revealed that RE-SPECT CVT has the lowest bias risk given utilisation of a Prospective Randomized Open, Blinded End-point (PROBE) design, and although the retrospective studies inherently have increased bias, most studies were appropriately informative. Finally, follow-up data and treatment duration were limited to a median 6 months; longer term registries for safety will be needed to estimate rates of recurrent CVT in patients treated with a DOAC.
Unanswered questions and future research
Our systematic review suggests physicians are increasingly using DOAC for the treatment of CVT; however, several remaining questions require further study. The ideal time to start a DOAC after diagnosis of CVT is not known. Certain studies first use LMWH or UFH treatment, while others used a DOAC acutely. The safety of DOAC use in children is not known. The recently published RCT, ‘Oral Rivaroxaban in Children With Venous Thrombosis’ (EINSTEIN-JR, NCT02234843), investigated paediatric cases of any acute VTE and randomised to weight-based rivaroxaban or standard anticoagulation showed potentially improved thrombotic burden (OR 1.70, p=0.012) and similar safety as adult studies.69 Specific outcomes were not reported based on VTE location; however, 74 of 335 (22%) patients treated with rivaroxaban had CVT, and no clear safety concern was identified. Finally, the ideal DOAC to use for CVT also requires further study. Results from RWCVT and SECRET will help validate safety and efficacy of rivaroxaban and allow more definitive comparison with dabigatran from RE-SPECT CVT.62 Although dabigatran has the advantage of having a reversal agent, idaricizumab, its use in CVT has not been published at the current time, so any unique risks in this population is unknown.70 Extrapolating conclusions for apixaban or edoxaban from studies of different DOAC may give an inaccurate risk–efficacy profile, and thus high-quality RCT of these treatments are also needed.
Given that CVT is a rare disease, enrolment in these large randomised studies is slow, so review of observational cohorts and smaller studies provide needed information. Physicians recognise the benefits of DOACs and are increasingly using these medications for treatment of CVT despite the lack of guideline recommendations. Based on this review, no clear safety concerns are identified for any particular DOAC, and the available data on efficacy is promising. The ideal timing for initiation of DOAC after diagnosis of CVT, and the ideal DOAC to use for CVT, are remaining questions. The results of future RCTs may inform guidelines if no adverse safety signal and a similar efficacy to standard therapy is seen.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors GB, JG, DD and RS developed the search strategy; GB, JG and DD reviewed articles for inclusion; GB, DAF and DD performed data analysis; VY assessed articles for risk of bias; GB wrote the manuscript; GLG, JC, MM, MV-B and SN contributed expert opinion and revised research question and discussion, and all authors revised the manuscript for intellectual content and approved the final manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests GB: none. JG: none. VY: none. RS: none. DAF: none. GLG holds an Early Researcher Award from the Ontario Ministry of Research and Innovation (MRI); an Ontario Mid-Career Investigator Award from the Heart and Stroke Foundation of Canada; and a University of Ottawa, Faculty of Medicine Tier 1 Clinical Research Chair in Diagnosis of Venous Thromboembolism. He has indirectly received research funding from Portola, Boehringer-Ingelheim, Pfizer, Bristol-Myers Squibb, LEO Pharma, Daiichi Sankyo, Bayer. He has received speaker honoraria from Bayer, Pfizer, LEO Pharma, Sanofi bioMérieux. JC: has received research funding from the following non-profit organisations: Dutch Heart Foundation, Dutch Brain Foundation and Amsterdam Neuroscience. He has also received research support from Bayer, Boehringer and Portola. All fees were paid to his institute and used to fund medical research. MM: none. MV-B received speaker fees from Boehringer Ingelheim, Portugal, is part of an advisory board of Daiichi Sankyo, Portugal, and received a travel grant from Boehringer Ingelheim, Portugal. SN: received consulting fees from Brainomix and Böhringer Ingelheim and Honoria for lectures from Bayer, BMS Pfizer and Medtronic. DD received a Heart & Stroke Foundation of Canada Clinician Scientist Award, and has received honoraria from Bayer, BMS and Apopharma.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available in a public, open access repository. Extra data can be accessed via the Dryad data repository at http://datadryad.org/ with the doi:10.5061/dryad.37pvmcvgn.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.