Cardiogenic embolism is a common cause of recurrent ischaemic stroke. The cardiac source of embolism is usually the left atrial appendage and atrium due to atrial fibrillation (AF). Other sources include the left ventricle, heart valves and venous system or right atrium, via a patent foramen ovale.The most effective thromboprophylactic is oral anticoagulation, which reduces the risk of recurrent stroke by about two thirds, compared to no anticoagulation. Four target-specific, direct-acting non-vitamin K antagonist oral anticoagulants (NOACs) – the direct thrombin inhibitor dabigatran etexilate, and the factor Xa inhibitors rivaroxaban, apixaban and edoxaban – are at least as efficacious and safe as warfarin, and apixaban is superior to, and as safe as, aspirin, for preventing stroke among patients with AF. Other potential stroke prevention strategies include left atrial appendage occlusion for patients with AF in whom anticoagulation is contraindicated, anticoagulation for left ventricular thrombus and prosthetic heart valves, antibiotics +/– valve surgery for infective endocarditis, and transcatheter device closure of a symptomatic patent foramen ovale.
It has traditionally been considered that about 20% of first-ever and recurrent ischaemic strokes are due to embolism from the heart. However, recent population-based studies, such as the 2009–2010 Adelaide Stroke Incidence Study, suggest that embolism from the heart is now the major cause of ischaemic stroke, accounting for 42% (95% confidence interval [CI]: 36–49%) of all ischaemic strokes in this community (Leyden et al., 2013).
The major cardiac sources of embolism are the left atrium and left atrial appendage (particularly atrial fibrillation [AF]), the left ventricle (myocardial infarction [MI], cardiomyopathy), the mitral and aortic valves (mitral stenosis, infective endocarditis), and the venous system and right atrium via a patent foramen ovale (paradoxical embolism).
This chapter will review the evidence for antithrombotic therapies, left atrial appendage occlusive devices, antibiotics, and percutaneous closure of a patent foramen ovale (PFO) to prevent recurrent ischaemic stroke due to embolism from the heart.
Atrial fibrillation is a major causal risk factor for embolic ischaemic stroke. It is conducive to thrombogenesis in the left atrium and left atrial appendage because it predisposes to each of Virchow’s classic triad: endothelial/endocardial damage or dysfunction, stasis of blood, and hypercoagulability of blood (Goldberger et al., 2015).
In view of the known association of AF with stroke, and the evidence that anticoagulation can prevent up to 80% of strokes in patients with clinical nonvalvular AF, it is attractive to hypothesize that screening high-risk, asymptomatic populations with electrocardiography (ECG) should identify patients with AF for whom anticoagulation can be initiated and thromboembolic complications such as ischaemic stroke prevented.
However, the US Preventive Services Task Force (USPSTF) has recently highlighted that, although ongoing studies are evaluating this hypothesis, there is no current evidence to support it. (US Preventive Services Task Force et al., 2018). Accordingly, the task force concludes that the current evidence is insufficient to assess the balance of benefits and harms of screening for AF with ECG.
This recommendation statement of the USPSTF is based on a recent evidence report and systematic review for the USPSTF, which concluded that although screening with ECG can detect previously unknown cases of AF, it has not been shown to detect more cases than screening focused on pulse palpation (Jonas et al., 2018). Furthermore, although treatments for AF reduce the risk of stroke and all-cause mortality, they increase the risk of bleeding. Trials have not assessed whether treatment of screen-detected asymptomatic older adults results in better health outcomes than treatment after detection by usual care or after symptoms develop.
A systematic review and random effects meta-analysis of 50 studies of the yield of cardiac monitoring for diagnosing new AF after stroke or transient ischaemic attack (TIA) in a total of 11,658 patients reported that the overall proportion of patients diagnosed with AF after stroke was
7.7% (95% CI: 5.0–10.8) by admission electrocardiogram (ECG) in the emergency room (phase 1);
5.1% (3.8–6.5) by serial ECG in hospital, continuous inpatient ECG monitoring, continuous inpatient cardiac telemetry, and in-hospital Holter monitoring (phase 2);
10.7% (5.6–17.2) by ambulatory Holter (phase 3); and
16.9% (13.0–21.2) by mobile cardiac outpatient telemetry, external loop recording, and implantable loop recording (phase 4) (Sposato et al., 2015).
The overall AF detection yield after all phases of sequential cardiac monitoring was 23.7% (95% CI: 17.2–31.0) (Sposato et al., 2015).
An analysis of individual patient data from three prospective studies of 1556 patients with stroke suggests that the AS5F score (Age: 0.76 points per year, Stroke severity National Institutes of Health Stroke Scale [NIHSS] <5 = 9 points, NIHSS >5 = 21 points) may be useful for selecting ischaemic stroke patients for prolonged ECG monitoring (to detect paroxysmal AF) (Uphaus et al., 2019). This score awaits external validation in independent data sets.
In the EMBRACE trial, noninvasive ambulatory ECG monitoring for 30 days in 572 patients with a recent cryptogenic stroke or TIA who were >55 years of age significantly improved the detection of AF by a factor of more than 5 (16.1% vs 3.2%, absolute difference 12.9%, 95% CI: 8.0–17.6%) as compared with the standard practice of short-duration 24-hour ECG monitoring. Noninvasive ambulatory ECG monitoring for 30 days also nearly doubled the rate of anticoagulant treatment at 90 days after randomization compared with the short-duration 24-hour ECG monitoring (18.6% vs 11.1%, absolute difference 7.5%, 95% CI: 1.6–13.3) (Gladstone et al., 2014).
In the CRYSTAL AF study, long-term monitoring with an insertable cardiac monitor (ICM) was more effective than conventional follow-up (control) for detecting AF at 6 months in 441 patients >40 years of age with cryptogenic stroke (<90 days ago) and no evidence of AF during >24 hours prior ECG monitoring (19 [8.9%] patients ICM vs 3 [1.4%] control; hazard ratio [HR] 6.4; 95% CI: 1.9–21.7; p < 0.001) (Sanna et al., 2014).
Atrial fibrillation might be newly detected in nearly a quarter of patients with stroke or TIA by sequentially combining cardiac monitoring methods.
Implications for Practice
It is recommended that for patients with acute ischaemic stroke or TIA and no apparent cause, prolonged rhythm monitoring (≈30 days) for AF is reasonable within 6 months of the event (Kernan et al., 2014).
Predictors of Stroke and Systemic Embolism in Patients with AF
For patients with AF, the main predictors of an increased risk of stroke are those comprising the CHADS2 and CHA2DS2-VASc scores (Tables 18.1 and 18.2) and the age, biomarkers (N-terminal fragment B-type natriuretic peptide [NT-proBNP] and cardiac troponin high sensitivity [cTn-hs]), and clinical history (prior stroke or TIA) (ABC) score (Lip et al., 2010; Olesen et al., 2011; Hijazi et al., 2016a; Borre et al., 2018; Chen et al., 2019). However, these scores demonstrate only mediocre predictive values with C statistics ranging from 0.55 to 0.64.
|Congestive heart failure||1||1|
|Age ≥75 years||1||2|
|Stroke, transient ischaemic attack, or thromboembolism||2||2|
|Vascular disease (previous MI, PAD, or aortic plaque)||1|
|Age 65–74 years||1|
|Sex category (female sex)||1|
LV: left ventricular. MI: myocardial infarction. PAD: peripheral artery disease.
|Prevalence||Stroke rate at 1 year,||Prevalence||Stroke rate at 1 year,|
|(%)||% (95% CI)||%||(95% CI)|
|0||22%||1.7% (1.5–1.9)||8%||0.8% (0.6–1.0)|
|1||31%||4.7% (4.4–5.1)||12%||2.0% (1.7–2.4)|
|2||23%||7.3% (6.9–7.8)||18%||3.7% (3.3–4.1)|
|3||15%||15.5% (14.6–16.3)||23%||5.9% (5.5–6.3)|
|4||7%||21.5% (20.0–23.2)||19%||9.3% (8.7–9.9)|
|5||2%||19.7% (16.9–22.9)||12%||15.3% (14.3–16.2)|
|6||0.2%||22.4% (14.6–34.3)||6%||19.7% (18.2–21.4)|
All patients in this study were discharged from a hospital in Denmark between 1997 and 2006 with non-valvular atrial fibrillation, were not anticoagulated, and were followed up for 1 year. CHA2DS2-VASc scores: 0 = no antithrombotic therapy recommended. 1 = antithrombotic therapy with oral anticoagulation or antiplatelet therapy recommended (preferably oral anticoagulation). ≥ 2 = oral anticoagulation recommended (Olesen et al., 2011).
Additional clinical predictors of stroke in anticoagulated patients with AF include vitamin K antagonist (VKA)-naive status (for VKA-experienced patients, risk ratio [RR] 0.85, 95% CI: 0.74–0.97), moderate renal impairment (1.54, 1.30–1.81), severe renal impairment (2.22, 1.85–2.66), previous aspirin use (1.19, 1.04–1.37), and Asian race (1.70, 1.42–2.03). (Albertsen et al., 2013).
Additional electrocardiographic predictors of ischaemic stroke in non-anticoagulated people with AF, independent of CHA2DS2-VASc variables, include an abnormal P-wave axis (Maheshwari et al., 2019).
Echocardiographic predictors of incident and recurrent stroke include echocardiographic evidence of stasis in the left atrium (spontaneous echo contrast [‘smoke’] or diminished left atrial appendage (LAA) flow velocities [specifically peak LAA emptying velocity <0.2 m/s]), large LA dimension/volume, and LAA geometry; one study reported that patients with chicken wing morphology of the LAA were associated with a lower risk of stroke or TIA compared to other LAA morphologies, such as the cactus, cauliflower, and windsock morphology (odds ratio [OR] 0.21, 95% CI: 0.05–0.91) (Di Biase et al., 2012; Goldberger et al., 2015).
An advantage of the CHA2DS2-VASc score over the CHADS2 score is that it helps identify very low-risk patients in whom anticoagulation may be associated with a net disadvantage (Friberg et al., 2012).
A limitation of the CHADS2 and CHA2DS2-VASc scores is that high scores are associated with increased risks of stroke and systemic embolism and also of bleeding and death (Oldgren et al., 2011). This association exists because many of the predictors of ischaemic stroke that contribute to the CHADS2 and CHA2DS2-VASc scores (see Table 18.1) are also predictors of major bleeding that comprise the HAS-BLED score (Table 18.3). One study showed that, of the so-called shared predictors, older age and previous stroke or TIA were more closely associated with ischaemic stroke than with intracerebral haemorrhage, whereas a history of hypertension, diabetes mellitus, renal impairment, and alcohol intake was not more strongly associated with either ischaemic or haemorrhagic stroke (McGrath et al., 2012).
The main predictors of an increased risk of major bleeding while taking oral anticoagulants are those comprising the HAS-BLED score (Tables 18.3 and 18.4) and ABC bleeding score (Age, Biomarkers [haemoglobin, cTn-hs, and GDF-15 or cystatin C/CKD-EPI] and Clinical history of previous bleeding]) (Pisters et al., 2010; Hijazi et al., 2016b; Borre et al., 2018).
|Hypertension (uncontrolled, systolic blood pressure ≥160 mm Hg)||1|
|Abnormal renal† or liver†† function (1 point each)||1 or 2|
|Stroke (previous history)||1|
|Bleeding history or predisposition (anaemia)||1|
|Labile INR (if on warfarin, time in the therapeutic range <60%)||1|
|Elderly (age >65 years)||1|
|Drugs (antiplatelet or NSAID)††† or excess alcohol†††† (1 point each)||1 or 2|
† Abnormal renal function is classified as the presence of chronic dialysis, renal transplantation, or serum creatinine ≥200 micromol/L.
†† Abnormal liver function is defined as chronic hepatic disease (e.g. cirrhosis) or biochemical evidence of significant hepatic derangement (bilirubin >2 times the upper limit of normal, in association with aspartate aminotransferase, alanine aminotransferase, or alkaline phosphatase >3 times the upper limit normal).
††† Drugs = concomitant use of antiplatelet or non-steroidal anti-inflammatory drugs (NSAIDs).
†††† Excess Alcohol = ≥8 units/week.
INR: international normalized ratio.
|HAS-BLED score||Patients, n||Bleeds, n||Bleeds per 100 patient-years|
Patients in the survey were followed up for 1 year. 48 patients had a major bleed at an overall rate of 1.56 per 100 patient-years. A HAS-BLED score of ≥3 indicates a major bleeding risk of about 3.7 per 100 patient-years, and means that caution is warranted in the prescription of oral anticoagulation, and treatment of modifiable risk factors for bleeding (e.g. hypertension, abnormal renal and liver function, drugs, and alcohol) and regular review are recommended.
In the Rivaroxaban Once Daily, Oral, Direct Factor Xa Inhibition Compared With Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF) trial cohort of 14,264 patients with AF who were anticoagulated with rivaroxaban or warfarin, 172 patients (1.2%) experienced 175 intracranial haemorrhage (ICH) events at a rate of 0.67% per year during 1.94 years (median) of follow-up (Hankey et al., 2014).
The significant, independent predictors of ICH were race (Asian: HR, 2.02; 95% CI: 1.39–2.94; black: HR, 3.25; 95% CI: 1.43–7.41), age (1.35; 1.13–1.63 per 10-year increase), reduced serum albumin (1.39; 1.12–1.73 per 0.5 g/dL decrease), reduced platelet count below 210 × 109/L (1.08; 1.02–1.13 per 10 × 109/L decrease), previous stroke or TIA (1.42; 1.02–1.96), and increased diastolic blood pressure (1.17; 1.01–1.36 per 10 mm Hg increase).
Predictors of a reduced risk of ICH were randomization to rivaroxaban (0.60; 0.44–0.82) and history of congestive heart failure (0.65; 0.47–0.89). The ability of the model to discriminate individuals with and without ICH was good (C-index, 0.69; 95% CI: 0.64–0.73).
Table 18.5 shows the platelets, albumin, no CHF, warfarin, age, race, diastolic blood pressure, stroke (PANWARDS) nomogram for predicting absolute risk of ICH in the ROCKET AF cohort of 14,000 anticoagulated patients with AF, which is based on the independent, significant prognostic factors for ICH and the strength of their association with risk of ICH.
|Platelets, × 109/L|
|No history of CHF||6|
|Warfarin instead of rivaroxaban||7|
|White or other||0|
|Diastolic blood pressure, mm Hg|
|Stroke or TIA in past||5|
CHF: congestive heart failure. PANWARDS: platelets, albumin, no CHF, warfarin, age, race, diastolic blood pressure, stroke. TIA: transient ischaemic attack.
Table 18.6 shows the predicted probabilities of ICH at 2.5 years according to the score derived from the variables within the PANWARDS nomogram.
|Total score||n||Probability of ICH at 2.5 years|
|n, number of patients|
Rivaroxaban Once Daily, Oral, Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF) cohort with PANWARDS score. ICH: intracranial haemorrhage. PANWARDS: platelets, albumin, no congestive heart failure, warfarin, age, race, diastolic blood pressure, stroke.
The ROCKET AF cohort did not measure or analyse all potential risk factors for ICH in anticoagulated AF patients, such as imaging features of leucoaraiosis and cerebral microbleeds, which could be incorporated in ICH risk scores (SPIRIT Trial Investigators, 1997; Wilson et al., 2018).
The CROMIS-2 observational study of 1490 patients with AF and recent acute ischaemic stroke or TIA, treated with a VKA or direct oral anticoagulant, and followed up for a mean period of 850 days (standard deviation [SD] 373) reported that the rate of symptomatic ICH in patients with cerebral microbleeds was 9.8 per 1000 patient-years (95% CI: 4.0–20.3) compared with 2.6 per 1000 patient-years (95% CI: 1.1–5.4) in those without cerebral microbleeds (adjusted HR: 3.67, 95% CI: 1.27–10.60) (Wilson et al., 2018).
Compared with the HAS-BLED score alone (C-index 0.41, 95% CI: 0.29–0.53), prediction models which included cerebral microbleeds and HAS-BLED (0.66, 0.53–0.80) and cerebral microbleeds, diabetes, anticoagulant type, and HAS-BLED (0.74, 0.60–0.88) predicted symptomatic ICH significantly better (difference in C-index 0.25, 95% CI: 0.07–0.43, p = 0.0065; and 0.33, 0.14–0.51, p = 0.00059, respectively) (Wilson et al., 2018).
The European Atrial Fibrillation Trial found that aspirin is not significantly more effective than control for preventing recurrent stroke in patients with ischaemic stroke or TIA with AF (HR 0.83, 95% CI: 0.65–1.05) (European Atrial Fibrillation Trial Study Group, 1993).
The result from this single secondary prevention trial in patients with prior stroke or TIA is consistent with the results of several primary prevention trials which also found that aspirin is not significantly more effective than placebo in preventing (first-ever) stroke in people with AF (RR 0.81, 95% CI: 0.65–1.01) (Hart et al., 2007).
The ACTIVE-A trial randomized 7554 patients with AF for whom VKA therapy was considered unsuitable (physician’s judgement that VKA was inappropriate, 50%; specific risk of bleeding, 23%; patient’s preference not to take VKA, 27%) to 75 mg clopidogrel once daily plus aspirin. Patients were followed up for a median of 3.6 years after randomization. Only 892 (13%) patients in this trial had a previous stroke or TIA (ACTIVE Investigators, 2009; Connolly et al., 2011a).
Dual antiplatelet therapy (75 mg clopidogrel once daily plus aspirin) was associated with a marginal but statistically significant reduction in the primary outcome of stroke, MI, non-central nervous system (CNS) systemic embolism, or death from vascular causes, after a median of 3.6 years of follow-up, compared with placebo plus aspirin (6.8% per year clopidogrel plus aspirin vs 7.6% per year aspirin; RR 0.89 [95% CI: 0.81–0.98]) (ACTIVE Investigators, 2009).
The combination of clopidogrel and aspirin was more effective than aspirin alone in preventing stroke (2.4% per year, clopidogrel plus aspirin, vs 3.3% per year, aspirin; RR 0.72, 95% CI: 0.62–0.83).
The ACTIVE-A trial data suggest that treating 1000 patients with AF for 1 year with clopidogrel plus aspirin prevents eight major vascular events (including two fatal and three disabling strokes) and causes seven major haemorrhages (one fatal) compared with aspirin alone.
Implications for Practice
For ischaemic stroke or TIA, AF, and being unable to take oral anticoagulants, the use of the combination of clopidogrel and aspirin therapy might be reasonable (Class IIb; Level of Evidence B) (Kernan et al., 2014).
Two randomized trials have compared oral anticoagulation versus control (no treatment) or placebo in a total of 485 patients with non-valvular AF and previous TIA or minor ischaemic stroke (Saxena and Koudstaal, 2004a).
For patients with AF and prior stroke or TIA, long-term adjusted-dose warfarin (international normalized ratio [INR]: 2.5–4.0) over about 2 years reduced the risk of recurrent stroke by two-thirds, compared with no antithrombotic therapy (8.9% warfarin vs 22.6% control, OR 0.36, 95% CI: 0.22–0.58) (Figure 18.1) (European Atrial Fibrillation Trial Study Group, 1993; Saxena and Koudstaal, 2004a)
Figure 18.1 Forest plot showing the effects of oral anticoagulants vs control in patients with nonrheumatic atrial fibrillation (NRAF) and a history of stroke or TIA on recurrent stroke at the end of follow-up.
Anticoagulation (INR 2.5 to 4.0 in the European Atrial Fibrillation Trial, 1993) significantly increased major extracranial bleeding complications compared with control (2.8% per year [oral anticoagulants] vs 0.7% per year [control]; OR 4.32, 95% CI: 1.55–12.1, absolute excess 2.1% per year) (Figure 18.2), but there was no association with any excess of intracranial bleeds compared to control in the two trials (OR 0.13, 95% CI: 0.00–6.49) (European Atrial Fibrillation Trial Study Group, 1993; Saxena and Koudstaal, 2004a).
Figure 18.2 Forest plot showing the effects of oral anticoagulants vs control in patients with NRAF and a history of stroke or TIA on major extracranial haemorrhages at the end of follow-up.
The lack of excess intracranial bleeding with warfarin is most likely explained by the inclusion of highly selected patients in the historical European Atrial Fibrillation trial and low statistical power to reliably identify or exclude a modest but clinically significant increase in risk.
The results of the two studies indicate that in patients with nonrheumatic atrial fibrillation (NRAF) and a recent TIA or minor ischaemic stroke, oral anticoagulant treatment decreases the odds of recurrent stroke (disabling and non-disabling) by two-thirds, from about 12% per year (control) to 4% per year (oral anticoagulants; absolute risk reduction [ARR] 8% per year) and almost halves the odds of serious vascular events.
Although there was a 4-fold increase in serious bleeding complications, the absolute excess of 2.1% per year did not negate the absolute benefit in preventing recurrent stroke (8% per year).
Implications for Practice
These results indicate that anticoagulants should be prescribed to patients with NRAF and a recent TIA or minor ischaemic stroke, unless there is a major contraindication.
The target value for the INR should be set between 2.0 and 3.0 for safe and effective stroke prevention.
Oral anticoagulation is generally started (or restarted) after about 2 weeks have elapsed because the risks of haemorrhagic transformation of the fresh brain infarct in the first 2 weeks may offset any benefits of earlier anticoagulation in reducing recurrent embolic ischaemic stroke. Patients with small ischaemic strokes, controlled blood pressure, and perhaps no evidence of microbleeding on gradient echo or susceptibility-weighted magnetic resonance imaging (MRI) sequences might be considered for earlier initiation of anticoagulant therapy, within the first week or so of ischaemic stroke, if they are at high risk of recurrent stroke. Hence, it is generally recommended that oral anticoagulation be initiated after the first 1–2 weeks of stroke onset. Patients with no residual brain infarction, such as those with TIA, should be safe to start anticoagulation immediately.
The Bridging Anticoagulation in Patients who Require Temporary Interruption of Warfarin Therapy for an Elective Invasive Procedure or Surgery (BRIDGE) was a randomized, double-blind, placebo-controlled trial which showed that in 1884 patients with AF who stopped warfarin therapy 5 days before an elective operation or other elective invasive procedure (and which was resumed within 24 hours after the procedure), random assignment to bridging anticoagulation therapy with low-molecular-weight heparin (100 International Units [IU] of dalteparin per kilogram of body weight) or matching placebo administered subcutaneously twice daily, from 3 days before the procedure until 24 hours before the procedure and then for 5 to 10 days after the procedure, forgoing bridging anticoagulation was noninferior to peri-operative bridging with low-molecular-weight heparin for the prevention of arterial thromboembolism and decreased the risk of major bleeding (Douketis et al., 2015).
The incidence of arterial thromboembolism (stroke, systemic embolism, or TIA) at 30 days after the procedure was 0.4% in the no-bridging group and 0.3% in the bridging group (risk difference, 0.1 %, 95% CI: –0.6–0.8; p = 0.01 for noninferiority). However, the incidence of major bleeding was 1.3% in the no-bridging group and 3.2% in the bridging group (relative risk, 0.41, 95% CI: 0.20–0.78; p = 0.005 for superiority) (Douketis et al., 2015).
Two randomized trials have compared oral anticoagulation versus single antiplatelet therapy in a total of 1371 patients with non-valvular AF and previous TIA or minor ischaemic stroke (Saxena and Koudstaal, 2004a).
The European Atrial Fibrillation Trial (EAFT) randomized 455 patients with AF and recent (<3 months) TIA or minor ischaemic stroke to receive either anticoagulants (INR 2.5 to 4.0) or aspirin (300 mg/day) for a mean follow-up period of 2.3 years.
The Studio Italiano Fibrillazione Atriale (SIFA) trial randomized 916 AF patients within 15 days of prior TIA or minor ischaemic stroke to anticoagulation (INR 2.0 to 3.5) or indobufen (a reversible platelet cyclooxygenase inhibitor, 100 or 200 mg bid) for a follow-up period of 1 year.
Adjusted-dose warfarin reduced the risk of recurrent stroke by about half (OR 0.49, 95% CI: 0.33–0.72) compared with antiplatelet therapy (European Atrial Fibrillation Trial Study Group, 1993; Saxena and Koudstaal, 2004b) (Figure 18.3).
Figure 18.3 Forest plot showing the effects of oral anticoagulants vs antiplatelet therapy in patients with NRAF and a history of stroke or TIA on recurrent stroke at the end of follow-up.
This proportional reduction is consistent with the effect of adjusted-dose warfarin compared with antiplatelet therapy in the prevention of first-ever stroke (RR 0.61 [0.48–0.78]) (Hart et al., 2007).
Anticoagulants were also significantly more effective than antiplatelet therapy in reducing the odds of all vascular events by about one-third (OR 0.67, 95% CI: 0.50–0.91).
Adjusted-dose warfarin increased major extracranial bleeding by several-fold compared with antiplatelet therapy (OR 5.2, 95% CI: 2.1–12.8) but the absolute difference observed in the trials was small (2.8% per year [oral anticoagulants] vs 0.9% per year [antiplatelet therapy] in EAFT and 0.9% per year [oral anticoagulants] vs 0% [antiplatelet therapy] in SIFA) (European Atrial Fibrillation Trial Study Group, 1993; Saxena and Koudstaal, 2004b) (Figure 18.4).
TIA and Ischaemic Stroke Patients
For individuals with AF and previous stroke or TIA, adjusted-dose warfarin was significantly more effective for prevention of stroke than the combination of clopidogrel plus aspirin (2.99% per year with warfarin vs 6.22% with clopidogrel plus aspirin; RR 0.47, 95% CI: 0.25–0.81) in the ACTIVE-W trial (Healey et al., 2008).
The benefit of oral anticoagulant over antiplatelet therapy in AF depended on the quality of INR control achieved by centres and countries as measured by time in therapeutic range (Connolly et al., 2008).
For patients enrolled in the ACTIVE-W trial with CHADS2 > 1, major bleeding occurred at a rate of 2.63% per year on clopidogrel plus aspirin and 2.75% per year on oral anticoagulation (RR 0.97, 95% CI: 0.69–1.35; p = 0.84). The relative risk of major bleeding with clopidogrel plus aspirin, compared with oral anticoagulation was not significantly different between patients with high and low CHADS2 scores (p for interaction = 0.15); however, the absolute risk of major bleeding on oral anticoagulation was significantly lower among patients with CHADS2 = 1 compared to CHADS2 > 1 (RR 0.49, 95% CI: 0.30–0.79; p = 0.0003).
Patients with ischaemic stroke/TIA and AF often have coexisting atherosclerotic vascular disease because hypertension and ischaemic heart disease are common causes of both AF and ischaemic stroke/TIA.
Traditionally, it has been considered that anticoagulation prevents the formation of fibrin-rich thrombus (so-called red clot) in areas of stasis of blood flow, such as the LAA associated with AF, whereas antiplatelet treatment prevents the formation of the platelet-rich thrombus (so-called white clot) in areas of high shear stress and endothelial injury associated with arterial vascular disease. Hence, for patients with AF who present with unstable vascular disease manifested by an acute coronary syndrome, or who are undergoing vascular injury by means of percutaneous coronary or carotid intervention or stenting, the combination of aspirin plus clopidogrel is usually prescribed, at least in the short term (to prevent arterial thrombosis), in addition to prophylactic anticoagulation (to prevent left atrial thrombosis).
However, the WOEST trial reported that among 573 anticoagulated patients with AF who had an acute coronary syndrome or underwent percutaneous coronary stent (drug eluting stent or bare metal stent) implantation, dual treatment with clopidogrel 75 mg daily plus oral anticoagulation (INR as originally indicated) caused less bleeding than triple therapy with clopidogrel 75 mg daily plus aspirin 80–100 mg daily plus oral anticoagulation (INR as originally indicated) (19.4% dual vs 44.4% triple; HR 0.36, 95% CI: 0.26–0.50) without increasing thrombotic events or the composite of death, MI, target vessel revascularization, stroke, or stent thrombosis (11.3% dual vs 17.7% triple; HR 0.60, 95% CI: 0.34–0.94) (DeWilde et al., 2013).
In view of the increased risk of bleeding with standard anticoagulation with a VKA plus DAPT with a P2Y12 inhibitor and aspirin in patients with AF undergoing percutaneous coronary intervention (PCI) with placement of stents, the PIONEER AF-PCI trial aimed to determine the effectiveness and safety of anticoagulation with a non-vitamin K antagonist oral anticoagulant (NOAC) (rivaroxaban) plus either one or two antiplatelet agents are (Gibson et al., 2016; Chi et al., 2018).
A total of 2124 participants with nonvalvular AF who had undergone PCI with stenting were randomly assigned to receive, in a 1:1:1 ratio, low-dose rivaroxaban (15 mg once daily) plus a P2Y12 inhibitor for 12 months (group 1); very-low-dose rivaroxaban (2.5 mg twice daily) plus DAPT for 1, 6, or 12 months (group 2); or standard therapy with a dose-adjusted VKA (once daily) plus DAPT for 1, 6, or 12 months (group 3).
The rates of clinically significant bleeding (the primary safety outcome) were lower in the two groups receiving rivaroxaban than in the group receiving standard therapy (16.8% in group 1, 18.0% in group 2, and 26.7% in group 3; hazard ratio for group 1 vs group 3, 0.59; 95% CI: 0.47–0.76; p < 0.001; hazard ratio for group 2 vs group 3, 0.63; 95% CI: 0.50–0.80; p < 0.001). The rates of death from cardiovascular causes, MI, or stroke were similar in the three groups (Kaplan–Meier estimates, 6.5% in group 1, 5.6% in group 2, and 6.0% in group 3; p-values for all comparisons were non-significant).
The RE-DUAL PCI trial randomly assigned 2725 patients with AF who had undergone PCI to triple therapy with warfarin plus a P2Y12 inhibitor (clopidogrel or ticagrelor) and aspirin (for 1 to 3 months) (triple-therapy group) or dual therapy with dabigatran (110 mg or 150 mg twice daily) plus a P2Y12 inhibitor (clopidogrel or ticagrelor) and no aspirin (110 mg and 150 mg dual-therapy groups). Outside the USA, elderly patients (≥80 years of age; ≥70 years of age in Japan) were randomly assigned to the 110 mg dual-therapy group or the triple-therapy group (Cannon et al., 2017).
The incidence of a major or clinically relevant nonmajor bleeding event (the primary endpoint) was 15.4% in the 110 mg dual-therapy group as compared with 26.9% in the triple-therapy group (HR 0.52; 95% CI: 0.42–0.63; p < 0.001 for noninferiority; p < 0.001 for superiority) and 20.2% in the 150 mg dual-therapy group as compared with 25.7% in the corresponding triple-therapy group, which did not include elderly patients outside the USA (HR 0.72; 95% CI: 0.58–0.88; p < 0.001 for noninferiority). The incidence of the composite efficacy endpoint was 13.7% in the two dual-therapy groups combined as compared with 13.4% in the triple-therapy group (HR 1.04; 95% CI: 0.84–1.29; p = 0.005 for noninferiority). The rate of serious adverse events did not differ significantly among the groups.
Interpretation of the Evidence
For patients with AF who have stable vascular disease, there is no reliable evidence that adding aspirin or clopidogrel, or both, to oral anticoagulation is safe and effective compared with anticoagulation alone.
For patients with AF who have unstable vascular disease, the WOEST trial suggests that adding aspirin to the combination of clopidogrel and effective anticoagulation may not prevent more ischaemic events yet may increase bleeding. The PIONEER AF PCI trial suggests that for patients with AF undergoing PCI, the administration of either low-dose rivaroxaban plus a P2Y12 inhibitor for 12 months or very-low-dose rivaroxaban plus DAPT for 1, 6, or 12 months is associated with a lower rate of clinically significant bleeding, and similar efficacy rates to standard therapy with a VKA plus DAPT for 1, 6, or 12 months. The RE-DUAL PCI trial suggests that for patients with AF who have undergone PCI, the risk of bleeding is lower among those who received dual therapy with dabigatran and a P2Y12 inhibitor than among those who received triple therapy with warfarin, a P2Y12 inhibitor, and aspirin. Dual therapy was also noninferior to triple therapy with respect to the risk of thromboembolic events.
A bivariate analysis that simultaneously characterized both risk and benefit demonstrated that rivaroxaban-based and dabigatran-based regimens were both favourable over VKA plus DAPT among patients with AF undergoing PCI (Chi et al., 2018).
Implication for Practice
Concomitant antiplatelet and anticoagulant therapy is not recommended in patients with AF and ischaemic stroke/TIA unless there is a specific indication, such as a mechanical heart valve (see below), or a recent acute coronary syndrome, or coronary stent, in which case the anticoagulant should probably be a NOAC rather than VKA.
Despite the effectiveness and affordability of vitamin K antagonists such as warfarin for the prevention of recurrent ischaemic stroke in patients with AF, these drugs have several limitations.
Warfarin has a slow onset of action and initially requires variable dose adjustments that may be due, in part, to genetic variability of vitamin K epoxide reductase complex 1 (which is involved in γ-carboxylation of vitamin K-dependent clotting factors) and cytochrome P450 enzymes, such as CYP2C9 (which are involved in the metabolism of warfarin).
Fluctuations in dietary intake of vitamin K and alcohol, and several drug–drug interactions, further contribute to inter-individual and intra-individual variability in anticoagulant effects of warfarin.
Consequently, despite close monitoring of coagulation to maintain the INR in the therapeutic range (INR 2.0–3.0), this ratio is frequently outside the therapeutic range, limiting the effectiveness of warfarin in up to 60% of patients with AF, and increasing the risk of bleeding (if INR too high) or thromboembolism (if INR too low).
The pharmacological characteristics of the NOACs are summarized in Table 18.7.
|Target||Thrombin||Factor Xa||Factor Xa||Factor Xa|
|Dosing||150 mg or 110 mg twice daily||15 or 20 mg once daily||2.5 or 5 mg twice daily||30 or 60 mg once daily|
|Time to peak plasma concentration||2 h||3 h||3 h||3 h|
|Half-life*||12–14 h||7–13 h||10–14 h||9–11 h|
|Renal excretion||80%||33% (66%)||25%||35%|
|Interactions||P-gp†||CYP450 3A4‡ P-gp†|
|Adverse effects||Dyspepsia, Bleeding||Bleeding||Bleeding||Bleeding|
* In patients with normal renal function.
† P-glycoprotein inhibitors include azole antifungals (e.g. ketoconazole, itraconazole, voriconazole, posaconazole) and protease inhibitors (e.g. ritonavir).
‡ Cytochrome P450 isoenzyme inhibitors include azole antifungals, protease inhibitors (e.g. atazanavir), and macrolide antibiotics (e.g. clarithromycin).
Dabigatran etexilate is a prodrug of dabigatran that directly inhibits thrombin. Thrombin has a pivotal role in blood coagulation by converting fibrinogen to fibrin; activating factors V, VIII, and XI; and activating platelets. Absorption from the gut depends on an acid environment, achieved with tartaric acid pellets coated with dabigatran etexilate. Despite this, bioavailability is only 6–7%. After oral administration, dabigatran etexilate is converted rapidly to dabigatran by hepatic and plasma esterases, and peak plasma concentrations are reached within 2 hours. Dabigatran is predominantly (80%) cleared by the kidneys. The drug half-life is 12–14 hours in patients with normal renal function, 18 hours if creatinine clearance is 30–50 mL/min, and more than 24 hours if creatinine clearance is less than 30 mL/min.
Dabigatran etexilate is a substrate for the P-glycoprotein transporter. P-glycoprotein is an efflux transporter that extrudes hydrophobic substances, such as toxins, out of cells into the gut, bile, and urine, and out of the brain and other organs. Since P-glycoproteins block absorption from the gut, co-administration of potent P-glycoprotein inhibitors (e.g. ketoconazole), which increase plasma concentrations of dabigatran, and co-administration of potent P-glycoprotein inducers (e.g. rifampicin), which reduce plasma concentrations of dabigatran, are contraindicated.
Rivaroxaban is a direct-acting factor Xa inhibitor with a half-life of 7–13 hours. The 20 mg dose has a bioavailability of about 66% in the fasted state and the 10 mg dose has a bioavailability of about 80–100%; co-administration with food increases the bioavailability. A third of rivaroxaban is excreted unchanged via the kidneys and the remainder is broken down via CYP3A4-dependent and CYP3A4-independent pathways in the liver and excreted as inactive metabolites in the urine (half) and faeces (half). Rivaroxaban is also a substrate for P-glycoprotein. Co-administration of potent inhibitors of both P-glycoprotein and CYP3A4 (e.g. ketoconazole) results in higher drug concentrations and is contraindicated.
Apixaban is a direct-acting factor Xa inhibitor with similar pharmacological properties to those of rivaroxaban. It is partly broken down via CYP3A4 and excreted via several pathways. A quarter is excreted via the kidneys and the half-life is about 12 hours. Co-administration of potent P-glycoprotein and CYP3A4 inhibitors is contraindicated because it can lead to apixaban toxicity with increased risk of bleeding.
Edoxaban is an oral, reversible, direct factor Xa inhibitor with a linear and predictable pharmacokinetic profile and 62% oral bioavailability. It achieves maximum concentrations within 1 to 2 hours, and 50% is excreted by the kidney. Pharmacokinetic modelling and simulation show that patients with low body weight, moderate-to-severe renal dysfunction, or concomitant use of a potent P-glycoprotein inhibitor should have the edoxaban dose reduced by 50%.
Advantages of the NOACs include a rapid onset of action; low propensity for interactions with food, alcohol, and drugs; and administration in fixed doses with a predictable anticoagulant effect that makes routine coagulation monitoring not necessary.
Despite concerns that measurement of drug concentration or anticoagulant activity might be needed to prevent excessively low or high drug concentrations, which may significantly increase thrombotic and bleeding risk, respectively, data from the ENGAGE AF-TIMI 48 trial suggest that adjustment of edoxaban dose based on clinical factors alone prevented excess drug concentration and the risk of bleeding events (Ruff et al., 2015). Edoxaban (or placebo-edoxaban in warfarin group) doses were halved at randomization or during the trial if patients had creatinine clearance 30–50 mL/min, body weight 60 kg or less, or concomitant medication with potent P-glycoprotein interaction. Although dose reduction decreased mean anti-FXa activity by 25% and 20% in the higher-dose and lower-dose regimens, respectively, dose reduction preserved the efficacy of edoxaban compared with warfarin (stroke or systemic embolism: higher dose p interaction = 0.85, lower dose p interaction = 0.99) and provided even greater safety (major bleeding: higher dose p interaction 0.02, lower dose p interaction = 0.002).
These findings validate the strategy that tailoring of the dose of edoxaban on the basis of clinical factors alone (renal function, body weight, concomitant P-glycoprotein transport inhibitor or inducer) achieves the dual goal of preventing excess drug concentrations and helping to optimize an individual patient’s risk of ischaemic and bleeding events. The results also showed that the therapeutic window for edoxaban is narrower for major bleeding than it is for thromboembolism.
Disadvantages include twice daily administration for dabigatran etexilate and apixaban; underdeveloped methods of coagulation monitoring; potential for drug toxicity in patients with renal insufficiency, liver insufficiency, or taking P glycoprotein or CYP3A4 inhibitors; and, until the US Food and Drug Administration (FDA) approved the use of idracuzimab as an antidote to dabigatran in 2015 and andexanet alfa as an antidote to rivaroxaban and apixaban in 2018, the absence of an antidote that is widely accessible and affordable, proven to reverse the anticoagulation effects of the drug, stop bleeding, and reduce death and disability due to bleeding (the latter of which remains to be proven).
The efficacy and safety of dabigatran, rivaroxaban, apixaban, and edoxaban have been compared with warfarin for stroke prevention in atrial fibrillation in four large, phase 3 clinical trials: RE-LY (Connolly et al., 2009, 2010), ROCKET AF (Patel et al., 2011), ARISTOTLE (Granger et al., 2011), and ENGAGE AF-TIMI (Giugliano et al., 2013; Ruff et al., 2014; Verheugt and Granger, 2015).
RE-LY (Randomised Evaluation of Long-term anticoagulant therapY) was a three-armed trial that compared two doses of dabigatran (110 mg twice a day or 150 mg twice a day) with standard dose-adjusted warfarin (target INR 2.0–3.0). Although healthcare providers and patients were blinded to dabigatran dose, warfarin was open label. All outcome events were independently adjudicated by two panel members blinded to treatment allocation.
Patients were excluded if they had poor renal function (creatinine clearance of <30 mL/min), active liver disease, or a stroke within 14 days of randomization, or were considered at high risk for bleeding. Twenty per cent of patients had a history of stroke or TIA. Low-dose aspirin was taken by about a third of all patients.
The lower dose of dabigatran etexilate (110 mg twice a day) was non-inferior to warfarin in reducing the rate of stroke or systemic embolism (warfarin 1.71% per year vs dabigatran 110 mg bid: 1.54% per year; RR 0.90, 95% CI: 0.74–1.10, p < 0.001 for non-inferiority, p = 0.30 for superiority).
The higher dose (150 mg twice a day) was superior to warfarin (warfarin: 1.71% per year vs dabigatran 150 mg bid: 1.11% per year; RR 0.65, 95% CI: 0.52–0.81; p < 0.001 for superiority).
The higher dose also significantly reduced the rate of ischaemic stroke compared with warfarin (RR 0.76, 95% CI: 0.59–0.97; p = 0.03).
The rates of all major haemorrhages were reduced in the lower-dose dabigatran group compared with warfarin (RR 0.80, 95% CI: 0.70–0.93), and roughly equal in the higher-dose dabigatran and warfarin groups (RR 0.93, 95% CI: 0.81–1.07). The exception was major gastrointestinal (GI) haemorrhage, which was increased in the higher-dose dabigatran group versus warfarin (1.56%/yr vs 1.15%/yr, RR 1.48, 95% CI: 1.18–1.85; p < 0.001).
Both doses of dabigatran significantly reduced haemorrhagic stroke and intracerebral haemorrhage compared with warfarin. The annual rates of haemorrhagic stroke were 0.12%/yr for the lower dose of dabigatran (HR vs warfarin 0.31, 95% CI: 0.17–0.56), 0.10%/yr for the higher dose (HR vs warfarin 0.26, 95% CI: 0.14–0.49), and 0.38%/yr for warfarin. The case fatality rate of intracerebral haemorrhage, which averaged 52%, was not significantly different in participants assigned dabigatran or warfarin (Hart et al., 2012).
Subsequent secondary analyses have shown that elderly patients (>75 years) had an increased risk of major haemorrhage with dabigatran 150 mg bid versus warfarin, and both doses of dabigatran had a higher risk of extracranial and GI haemorrhage compared with warfarin. However, the rate of ICH remained less with dabigatran than warfarin for all age groups.
Dabigatran caused higher rates of dyspepsia than did warfarin (11.8% with dabigatran 110 mg twice a day and 11.3% with dabigatran 150 mg twice a day vs 5.8% with warfarin), presumably related to the tartaric acid content of the dabigatran etexilate capsule.
A numerical increase in MI was reported with both doses of dabigatran compared with warfarin. A meta-analysis of all trials of dabigatran, including RE-LY, suggests that the absolute increase in MI is slight (0.14–0.17% per year), not statistically significant, and outweighed by the reduction in stroke and systemic embolism (0.6% per year).
The relative effects of dabigatran versus warfarin in the 3623 patients with previous stroke or TIA were consistent with the effects of dabigatran versus warfarin in the 14,490 patients without previous stroke or TIA for both stroke or systemic embolism and major bleeding (Diener et al., 2010).
The rate of stroke or systemic embolism among patients treated with dabigatran 110 mg per day compared with warfarin was consistent among patients with prior stroke or TIA (2.32%/year dabigatran 110 mg vs 2.78%/year warfarin; RR 0.84, 95% CI: 0.58–1.20) and patients without prior stroke or TIA (1.34%/year vs 1.45%/year; RR 0.93, 95% CI: 0.73–1.18; interaction p = 0.62).
The rate of stroke or systemic embolism among patients treated with dabigatran 150 mg per day compared with warfarin was consistent among patients with prior stroke or TIA (2.07%/year dabigatran 150 mg vs 2.78%/year warfarin; RR 0.75, 95% CI: 0.52–1.08) and patients without prior stroke or TIA (0.87%/year vs 1.45%/year; RR 0.60, 95% CI: 0.45–0.78; interaction p = 0.34).
The rate of major bleeding among patients treated with dabigatran 110 mg per day compared with warfarin was consistent among patients with prior stroke or TIA (2.74%/year dabigatran 110 mg vs 4.15%/year warfarin; RR 0.66, 95% CI: 0.48–0.90) and patients without prior stroke or TIA (2.91%/year vs 3.43%/year; RR 0.85, 95% CI: 0.72–0.99; interaction p = 0.15).
The rate of major bleeding among patients treated with dabigatran 150 mg per day compared with warfarin was consistent among patients with prior stroke or TIA (4.15%/year dabigatran 150 mg vs 4.15%/year warfarin; RR 1.01, 95% CI: 0.77–1.34) and patients without prior stroke or TIA (3.10%/year vs 3.43%/year; RR 0.91, 95% CI: 0.77–1.06; interaction p = 0.51).
The oral factor Xa inhibitor rivaroxaban was compared with warfarin in ROCKET AF (Rivaroxaban – Once daily oral direct factor Xa inhibition compared with vitamin K antagonism [target INR 2.0–3.0] for prevention of stroke and Embolism Trial in Atrial Fibrillation) (Patel et al., 2011). The study population in ROCKET AF was at high risk of stroke; 55% of patients had a previous stroke or TIA, and 90% had either a previous stroke or TIA, or three or more risk factors for stroke. Patients were randomly assigned to receive fixed-dose rivaroxaban (20 mg daily, or 15 mg daily in patients with a creatinine clearance of 30–49 mL per minute) or adjusted-dose warfarin (target INR 2.0–3.0).
Rivaroxaban was non-inferior to warfarin for the prevention of stroke or systemic embolism in the primary per-protocol, on-treatment analysis (1.71% per year rivaroxaban vs 2.16% per year warfarin, HR: 0.79; 95% CI: 0.66–0.96; p < 0.001 for noninferiority), but was not better than warfarin according to the intention-to-treat analysis (2.12% vs 2.42% per year; HR 0.88; 95% CI: 0.74–1.03; p < 0.001 for noninferiority; p = 0.117 for superiority).
Bleeding – The rates of major and clinically relevant non-major bleeding were similar with rivaroxaban (14.91% per year) and warfarin (14.52% per year; HR 1.03, 95% CI: 0.96–1.11; p = 0.44).
However, rivaroxaban was associated with lower rates of intracranial haemorrhage (p = 0.019) and fatal bleeding (0.24% per year with rivaroxaban vs 0.48% with warfarin; HR 0.50, 95% CI: 0.31–0.79), but higher rates of major gastrointestinal bleeding (3.15% vs 2.16% per year; p < 0.001).
The relative effects of rivaroxaban versus warfarin in the 7468 patients with previous stroke or TIA were consistent with the effects of rivaroxaban versus warfarin in the 6796 patients without previous stroke or TIA for stroke or systemic embolism (Hankey et al., 2012).
Stroke or Systemic Embolism
The rate of stroke or systemic embolism among patients treated with rivaroxaban compared with warfarin was consistent among the 7468 patients with prior stroke or TIA (2.79%/year rivaroxaban vs 2.96%/year warfarin; HR 0.94, 95% CI: 0.77 to 1.16) and the 6796 patients without prior stroke or TIA (1.44%/year vs 1.88%/year; HR 0.77, 95% CI: 0.58–1.01; interaction p = 0.23).
The rate of major bleeding among patients treated with rivaroxaban compared with warfarin was also consistent among patients with prior stroke or TIA (3.13%/year rivaroxaban vs 3.22%/year warfarin; HR 0.97, 95% CI: 0.79–1.19) and patients without prior stroke or TIA (4.10%/year vs 3.69%/year; HR 1.11, 95% CI: 0.92–1.34; interaction p = 0.36).
The ARISTOTLE (Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation) trial compared apixaban with warfarin (target INR 2.0–3.0) (Granger et al., 2011). Patients with impaired renal function (serum creatinine >2.5 mg/dL [221 μmol/L] or creatinine clearance <25 mL/min) were excluded. Patients who were elderly (80 years or older), had low body weight (60 kg or lighter), or a serum creatinine of 133 μmol/L (1.5 mg/dL) or greater received a lower dose of apixaban (2.5 mg twice a day) than did other patients (5 mg twice a day).
Apixaban was better than warfarin in reducing the rate of stroke or systemic embolism (1.27% per year apixaban vs 1.60% per year warfarin; HR: 0.79, 95% CI: 0.66–0.95; p < 0.001 for noninferiority; p = 0.01 for superiority).
Apixaban was also associated with significantly less major bleeding (2.13% per year apixaban vs 3.09% per year warfarin; HR 0.69, 95% CI: 0.60–0.80; p < 0.001), less ICH (0.33% per year apixaban vs 0.80% per year warfarin; HR 0.42, 95% CI: 0.30–0.58; p < 0.001), and lower mortality (3.52% apixaban vs 3.94% warfarin; HR 0.89, 95% CI: 0.80–0.99; p = 0.047) than warfarin.
The occurrence of gastrointestinal haemorrhage and MI was similar in both groups.
The relative effects of apixaban versus warfarin in the 3436 patients with previous stroke or TIA in the ARISTOTLE trial were consistent with the effects of apixaban versus warfarin in the 14,765 patients without previous stroke or TIA for stroke or systemic embolism and major bleeding (Diener et al., 2012).
The rate of stroke or systemic embolism among patients treated with apixaban compared with warfarin was consistent among the 3436 patients with prior stroke or TIA (2.46%/year apixaban vs 3.24%/year warfarin; HR 0.76, 95% CI: 0.56–1.03) and the 14,765 patients without prior stroke or TIA (1.01%/year vs 1.23%/year; HR 0.82, 95% CI: 0.65–1.03; interaction p = 0.71).
The rate of major bleeding among patients treated with apixaban compared with warfarin was also consistent among patients with prior stroke or TIA (2.84%/year apixaban vs 3.91%/year warfarin; HR 0.73, 95% CI: 0.55–0.98) and patients without prior stroke or TIA (1.98%/year vs 2.91%/year; HR 0.68, 95% CI: 0.58–0.80; interaction p = 0.69).
The Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation–Thrombolysis in Myocardial Infarction 48 (ENGAGE AF-TIMI 48) trial was a three-group, randomized, double-blind, double-dummy trial comparing two once-daily regimens of edoxaban (higher-dose edoxaban [60 mg once daily], or lower-dose edoxaban [30 mg once daily]) with adjusted dose warfarin (INR 2.0–3.0) in 21,105 patients with moderate-to-high risk of AF (median follow-up, 2.8 years) (Giugliano et al., 2013).
The ENGAGE AF-TIMI 48 trial showed that both once-daily regimens of edoxaban were noninferior to warfarin with respect to the prevention of stroke or systemic embolism and were associated with significantly lower rates of bleeding and death from cardiovascular causes.
The annualized rate of stroke or systemic embolism during treatment was 1.50% with warfarin (median time in the therapeutic range, 68.4%), as compared with 1.18% with high-dose edoxaban (HR 0.79; 97.5% CI: 0.63–0.99; p < 0.001 for noninferiority) and 1.61% with low-dose edoxaban (HR 1.07; 97.5% CI: 0.87–1.31; p = 0.005 for noninferiority).
In the intention-to-treat analysis, there was a trend favouring high-dose edoxaban versus warfarin (HR 0.87; 97.5% CI: 0.73–1.04; p = 0.08) and an unfavourable trend with low-dose edoxaban versus warfarin (HR 1.13; 97.5% CI: 0.96–1.34; p = 0.10).
Patients randomized to high-dose edoxaban had fewer strokes on-treatment (HR 0.80; 95% CI: 0.65–0.98) than those on warfarin (median time-in-therapeutic range, 68.4%); patients in the low-dose edoxaban group had similar rates (HR 1.10 versus warfarin; 95% CI: 0.91–1.32) (Giugliano et al., 2014).
Rates of ischaemic stroke or TIA were similar with high-dose edoxaban (1.76% per year) and warfarin (1.73% per year; P = 0.81), but more frequent with low-dose edoxaban (2.48% per year; p < 0.001).
The annualized rate of major bleeding was 3.43% with warfarin versus 2.75% with high-dose edoxaban (HR 0.80; 95% CI: 0.71–0.91; p < 0.001) and 1.61% with low-dose edoxaban (HR 0.47; 95% CI: 0.41–0.55; p < 0.001).
Both edoxaban regimens significantly reduced haemorrhagic stroke and other subtypes of intracranial bleeds.
Other Outcome Events
The corresponding annualized rates of death from cardiovascular causes were 3.17% versus 2.74% (HR 0.86; 95% CI: 0.77–0.97; p = 0.01), and 2.71% (HR 0.85; 95% CI: 0.76–0.96; p = 0.008), and the corresponding rates of the composite of stroke, systemic embolism, or death from cardiovascular causes were 4.43% versus 3.85% (HR 0.87; 95% CI: 0.78–0.96; p = 0.005), and 4.23% (HR 0.95; 95% CI: 0.86–1.05; p = 0.32).
Atrial Fibrillation Patients
A meta-analysis, using a random effects model, of all 71,683 participants included in the RE-LY, ROCKET AF, ARISTOTLE, and ENGAGE AF-TIMI 48 trials, reported that a total of 42,411 participants received a new oral anticoagulant and 29,272 participants received warfarin (Ruff et al., 2014).
The meta-analysis showed that the NOACs have a favourable risk–benefit profile, with significant reductions in stroke, ICH, and mortality, and with similar major bleeding to warfarin but with increased gastrointestinal bleeding. The relative efficacy and safety of new oral anticoagulants are consistent across a wide range of patients. Updated meta-analyses support these observations (Lowenstern et al., 2018).
The NOACs significantly reduced stroke or systemic embolic events by 19% compared with warfarin (RR 0.81, 95% CI: 0.73–0.91; p < 0.0001; I2 = 47%, heterogeneity p = 0.13). This was mainly driven by a reduction in haemorrhagic stroke (0.49, 0.38–0.64; p < 0.0001). There was no heterogeneity for stroke or systemic embolic events in important clinical subgroups.
Although indirect comparisons suggest that the relative effects of each of the NOACs compared with warfarin were reasonably consistent for stroke and systemic embolism (p-value for heterogeneity = 0.13), indirect comparisons of the NOACs should be interpreted cautiously because the trials had different designs, participants, and interventions (e.g. time in the therapeutic range [TTR] in those assigned warfarin differed between the trials).
Low-dose NOAC regimens showed similar overall reductions in stroke or systemic embolic events compared with warfarin (RR: 1.03, 0.84–1.27; p = 0.74), but significantly more ischaemic strokes (RR 1.28, 1.02–1.60; p = 0.045).
The NOACs were associated with a non-significant trend towards reduced major bleeding events by 14% compared with warfarin, but there was substantial heterogeneity among the trials (RR 0.86, 95% CI: 0.73–1.00; p = 0.06; I2 = 83%, heterogeneity p = 0.001). There was a greater relative reduction in major bleeding with NOACs when the centre-based time in the therapeutic range was less than 66% than when it was 66% or more (RR 0.69, 95% CI: 0.59–0.81 vs RR 0.93, 95% CI: 0.76–1.13; p for interaction = 0.022).
Low-dose NOAC regimens showed a trend towards a more favourable bleeding profile compared with warfarin (RR 0.65, 95% CI: 0.43–1.00; p = 0.05) (Huang et al., 2018).
Other Outcome Events
The NOACs also significantly reduced all-cause mortality (RR 0.90, 95% CI: 0.85–0.95; p = 0.0003) and ICH (RR 0.48, 95% CI: 0.39–0.59; p < 0.0001), but increased gastrointestinal bleeding (RR 1.25, 95% CI: 1.01–1.55; p = 0.04).
Low-dose new oral anticoagulant regimens showed similar overall rates of stroke or systemic embolic events compared with warfarin (RR 1.03, 0.84–1.27; p = 0.74) but significantly more ischaemic strokes (RR 1.28, 1.02–1.60; p = 0.045).
The relative effects of the NOACs versus warfarin in preventing stroke or systemic embolism were consistent among patients with AF and previous TIA or ischaemic stroke (4.94% NOAC vs 5.73% warfarin; RR 0.86; 95% CI: 0.76–0.98) and among patients with AF and no history of previous TIA or ischaemic stroke (2.33% NOAC vs 2.98% warfarin; RR 0.78, 95% CI: 0.66–0.91; interaction p = 0.30) (Ruff et al., 2014; Ntaios et al., 2017).
The relative effects of the NOACs vs warfarin on major bleeding were consistent among patients with AF and previous TIA or ischaemic stroke (5.71% NOAC vs 6.43% warfarin; RR 0.89, 95% CI: 0.77–1.02) and among patients with AF and no history of previous TIA or ischaemic stroke (5.18% NOAC vs 6.21% warfarin; RR 0.85; 95% CI: 0.72–1.01); interaction p = 0.70 (Ruff et al., 2014; Ntaios et al., 2017).
A systematic review identified eight randomized controlled trials (RCTs) comparing DTIs versus VKAs for prevention of stroke and systemic embolism in a total of 27,557 participants with non-valvular AF and one or more risk factors for stroke (Salazar et al., 2014).
The DTIs included dabigatran 110 mg twice daily and 150 mg twice daily (three studies, 12,355 participants), AZD0837 300 mg once per day (two studies, 233 participants), and ximelagatran 36 mg twice per day (three studies, 3726 participants). The VKA comparator was warfarin (10,287 participants).
The review showed that DTIs were as efficacious as VKAs for the composite outcome of vascular death and ischaemic events, and only the dose of dabigatran 150 mg twice daily was found to be superior to warfarin. DTIs were associated with fewer major haemorrhagic events, including haemorrhagic strokes. Adverse events that led to discontinuation of treatment occurred more frequently with the DTIs. There was no difference in death from all causes.
The odds of vascular death and ischaemic events were not significantly different between all DTIs and warfarin (OR 0.94, 95% CI: 0.85–1.05) (Figure 18.5).
Sensitivity analysis by dose of dabigatran on reduction in ischaemic events and vascular mortality indicated that dabigatran 150 mg twice daily was superior to warfarin, although the effect estimate was of borderline statistical significance (OR 0.86, 95% CI: 0.75–0.99). Sensitivity analyses by other factors did not alter the results.
Figure 18.5 Forest plot showing the effects of direct thrombin inhibitors vs vitamin K antagonists in patients with non-valvular atrial fibrillation on ischaemic events and vascular death at the end of follow-up.
Fatal and non-fatal major bleeding events, including haemorrhagic strokes, were less frequent with the DTIs (OR 0.87, 95% CI: 0.78–0.97) (Figure 18.6).
Figure 18.6 Forest plot showing the effects of direct thrombin inhibitors vs vitamin K antagonists in patients with non-valvular atrial fibrillation on fatal and non-fatal haemorrhage at the end of follow-up.
Other Outcome Events
Adverse events that led to discontinuation of treatment were significantly more frequent with the DTIs (OR 2.18, 95% CI: 1.82–2.61). All-cause mortality was similar between DTIs and warfarin (OR 0.91, 95% CI: 0.83–1.01).
Atrial Fibrillation Patients
A systematic review identified 13 RCTs that directly compared the effects of long-term treatment (>4 weeks) with factor Xa inhibitors and a VKA (warfarin) for the prevention of stroke and systemic embolism in 67,688 participants with a confirmed diagnosis of AF (or atrial flutter) (Bruins Slot and Berge, 2018).
The trials directly compared dose-adjusted warfarin, with a target INR of 2.0 to 3.0, with either apixaban, betrixaban, darexaban, edoxaban, idraparinux, idrabiotaparinux, or rivaroxaban. The majority of the included data (approximately 90%) were from apixaban, edoxaban, and rivaroxaban.
The review found that factor Xa inhibitors significantly reduced the number of strokes and systemic embolic events compared with warfarin in patients with AF. The absolute effect of factor Xa inhibitors compared with warfarin treatment was, however, rather small.
Factor Xa inhibitors also reduced the number of ICHs, all-cause deaths, and major bleedings compared with warfarin, although the evidence for a reduction of major bleedings is less robust.
The composite primary efficacy endpoint of all strokes (both ischaemic and haemorrhagic) and non-central nervous systemic embolic events was reported in all of the included studies.
Treatment with a factor Xa inhibitor significantly decreased the composite of stroke and systemic embolic events compared with dose-adjusted warfarin in participants with AF (OR 0.89, 95% CI: 0.82–0.97; 13 studies; 67,477 participants; high-quality evidence) (Figure 18.7).