Challenges in stroke prevention among very elderly patients with atrial fibrillation: discerning facts from prejudices

This editorial refers to ‘Reduced dose of rivaroxaban and dabigatran vs. vitamin K antagonists in very elderly patients with atrial fibrillation in a nationwide cohort study’, by L. Fauchier et al., on pages 205--215. The lifetime risk for atrial fibrillation (AF) by age of 90 has been recently estimated to 1 in 3 individuals of European ancestry, with the number of AF patients in the European Union reaching nearly 15 million by 2060 and >65% of AF patients being older than 80 years.1 Age is a prominent risk factor for many non-communicable diseases including AF, but the health and functional status of an elderly individual depends on multiple factors and may be only loosely associated with calendar age. Indeed, inevitable age-related decline in certain biological functions (Figure 1, upper panel) can be substantially delayed or compensated by healthy behaviours and healthcare environments supporting the concept of healthy aging, also including timely diagnosis and management of pathological conditions. Thus, ‘biological age’ could be more relevant than ‘chronological age’ even assessing elderly patients. Nevertheless, very old patients (e.g. those aged >85–90 years) are often undertreated owing to the (mis)perception of their needs and net benefit from specific treatments, as well as the healthcare system-related barriers. Factors influencing individual aging and health characteristics in the elderly (upper panel). Proportions of older patients in randomized clinical trials of OAC for stroke prevention in AF patients and NOAC dose reduction criteria (lower panel). aNumber of patients aged ≥85 years not reported, but 41.3% of patients were ≥80 years old. AF, atrial fibrillation; ARISTOTLE, Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation; AVERROES, Apixaban Versus Acetylsalicylic Acid [ASA] to Prevent Stroke in Atrial Fibrillation Patients Who Have Failed or Are Unsuitable for Vitamin K Antagonist Treatment; BAFTA, Birmingham Atrial Fibrillation Treatment of the Aged; bid, bis in die (twice daily); COPD, chronic obstructive pulmonary disease; CrCl, creatinine clearance; ENGAGE AF, Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation; OAC, oral anticoagulant therapy; od, once daily; QoL, quality of life; RE-LY, Randomized Evaluation of Long-Term Anticoagulation Therapy; ROCKET AF, Rivaroxaban Once Daily Oral Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation; sCr, serum creatinine. Age is an independent risk factor for adverse outcomes in AF patients including stroke—a devastating complication of AF that can be effectively prevented using oral anticoagulant therapy (OAC) with non-vitamin K antagonist oral anticoagulants (NOACs) or well-managed vitamin K antagonists (VKAs), the former being increasingly preferred as a more convenient and safer option.2 Observational data also suggest that OAC (especially NOACs) may provide additional protection from dementia,3,4 and several ongoing randomized clinical trials (RCTs) are investigating these observations. Nevertheless, older AF patients are often denied OAC and prescribed aspirin (or even no antithrombotic therapy) for stroke prevention owing to concerns about aging-related health characteristics such as frailty, cognitive impairment/dementia or multimorbidity, and the risk of OAC-related bleeding (Figure 1, upper panel). Indeed, age is an independent risk factor for OAC-related bleeding,5 but in clinical practice bleeding risk is often overrated in the elderly despite sufficient evidence from RCTs,2,6 meta-analyses,7 or large-scale observational datasets8 showing that the benefits from OAC (especially NOACs) outweigh the risk of bleeding in AF patients aged ≥75 years, whereas aspirin is less effective and offers no safety advantage (or may even be harmful).9 A recent retrospective analysis of a large administrative dataset showed consistent performance of NOACs compared with VKAs among AF patients with or without geriatric characteristics such as age ≥75 years, multimorbidity, polypharmacy, or frailty,8 whilst another analysis suggested that even in older patients (mean age 81 years) with contraindications to OAC [predominantly a blood dyscrasia, prior intracranial haemorrhage (ICH) or major gastrointestinal bleeding]. OAC use was associated with reduced risk for all-cause death, any stroke, or all-cause hospitalization, but at the cost of a 42% increased risk for ICH.10 Since the survivors of a trauma-related ICH would likely benefit from OAC, especially NOACs, in contrast to those having a small cerebral vessel disease or microbleeds,6 these findings emphasize the need for highly individualized approaches in the risk–benefit assessment among such complex AF patients. Whereas some patients aged ≥75 years were represented in the RCTs of OAC for stroke prevention in AF, those very old (≥85 years) comprised only around 5% of the studied patients in the landmark NOACs trials (Figure 1, lower panel). Notably, OAC is still significantly underused among such patients in clinical practice and, given the globally increasing proportion of very old patients with AF,1,9 their evidence-based management remains a major unmet need. What do we know about antithrombotic treatments for stroke prevention in very old AF patients? A recent nationwide AF cohort study from Taiwan9 reported reassuring effectiveness and safety of VKAs vs. no therapy (or aspirin) and NOACs vs. VKAs among very old AF patients (see Table 1), consistent with a few previously published registry-based or single centre observational studies, thus supporting the use of OAC (preferably NOACs) in very old AF patients. Major findings from ‘real-world’ atrial fibrillation cohorts of very old patients with atrial fibrillation 4.4% in the ‘VKA era’ (1996–2011) 8.6% in the ‘NOAC era’ (2012–15) 3.9% in the ‘VKA era’ 16.1% in the ‘NOAC era’ VKAs vs. no therapy Stroke risk reduction with no significant difference in major bleeding/ICH Ischaemic stroke 3.83%/year vs. 7.01%/year HR 0.61, 0.40–0.94 ICH HR 1.46, 0.57–3.71 NOACs vs. warfarin—effectiveness Similar or better effectiveness of NOACs in stroke risk reduction as compared to VKAs Ischaemic stroke 4.07%/year vs. 4.59%/year HR 1.16, 0.61–2.22 Stroke/SE Rivaroxaban 15 mg vs. VKA 3.8%/year vs. 3.3%/year HR 1.14, 0.93–1.40 Dabigatran 110 mg vs. VKA 2.8% vs. 3.6% HR 0.77, 0.60–0.99 NOACs vs. warfarin—safety Major bleeding Similar safety regarding major bleeding 6.14%/year vs. 6.75%/year HR 0.95, 0.63–1.44 Rivaroxaban 15 mg vs. VKA 3.7%/year vs. 4.0%/year HR 0.91, 0.74–1.11 Dabigatran 110 mg vs. VKA 3.3% vs. 4.0% HR 0.81, 0.64–1.03 NOACs vs. warfarin—safety ICH Significantly greater safety regarding ICH 0.42%/year vs. 1.63%/year HR 0.32, 0.10–0.97 Rivaroxaban 15 mg vs. VKA 0.8%/year vs. 1.1%/year HR 0.80, 0.54–1.20 Dabigatran 110 mg vs. VKA 0.5%/year vs. 1.2%/year HR 0.38, 0.22–0.66 NOACs vs. warfarin—safety Gastrointestinal bleeding Similar or greater risk for GI major bleeding with NOACs compared to VKAs Rivaroxaban 15 mg vs. VKA 2.5%/year vs. 2.0%/year HR 1.27, 0.98–1.65 Dabigatran 110 mg vs. VKA 3.2%/year vs. 1.7%/year HR 1.87, 1.40–2.51 All-cause mortality Similar or lower all-cause mortality with NOACs in comparison to VKAs Rivaroxaban 15 mg vs. VKA 17.5%/year vs. 19.1%/year HR 0.91, 0.83–1.00 Dabigatran 110 mg vs. VKA 16.4%/year vs. 18.7%/year HR 0.87, 0.78–0.97 Warfarin vs. no therapy POSITIVE NCB Warfarin vs. aspirin POSITIVE NCB Rivaroxaban 15 mg vs. VKA Neutral Dabigatran 110 mg vs. VKA Positive 4.4% in the ‘VKA era’ (1996–2011) 8.6% in the ‘NOAC era’ (2012–15) 3.9% in the ‘VKA era’ 16.1% in the ‘NOAC era’ VKAs vs. no therapy Stroke risk reduction with no significant difference in major bleeding/ICH Ischaemic stroke 3.83%/year vs. 7.01%/year HR 0.61, 0.40–0.94 ICH HR 1.46, 0.57–3.71 NOACs vs. warfarin—effectiveness Similar or better effectiveness of NOACs in stroke risk reduction as compared to VKAs Ischaemic stroke 4.07%/year vs. 4.59%/year HR 1.16, 0.61–2.22 Stroke/SE Rivaroxaban 15 mg vs. VKA 3.8%/year vs. 3.3%/year HR 1.14, 0.93–1.40 Dabigatran 110 mg vs. VKA 2.8% vs. 3.6% HR 0.77, 0.60–0.99 NOACs vs. warfarin—safety Major bleeding Similar safety regarding major bleeding 6.14%/year vs. 6.75%/year HR 0.95, 0.63–1.44 Rivaroxaban 15 mg vs. VKA 3.7%/year vs. 4.0%/year HR 0.91, 0.74–1.11 Dabigatran 110 mg vs. VKA 3.3% vs. 4.0% HR 0.81, 0.64–1.03 NOACs vs. warfarin—safety ICH Significantly greater safety regarding ICH 0.42%/year vs. 1.63%/year HR 0.32, 0.10–0.97 Rivaroxaban 15 mg vs. VKA 0.8%/year vs. 1.1%/year HR 0.80, 0.54–1.20 Dabigatran 110 mg vs. VKA 0.5%/year vs. 1.2%/year HR 0.38, 0.22–0.66 NOACs vs. warfarin—safety Gastrointestinal bleeding Similar or greater risk for GI major bleeding with NOACs compared to VKAs Rivaroxaban 15 mg vs. VKA 2.5%/year vs. 2.0%/year HR 1.27, 0.98–1.65 Dabigatran 110 mg vs. VKA 3.2%/year vs. 1.7%/year HR 1.87, 1.40–2.51 All-cause mortality Similar or lower all-cause mortality with NOACs in comparison to VKAs Rivaroxaban 15 mg vs. VKA 17.5%/year vs. 19.1%/year HR 0.91, 0.83–1.00 Dabigatran 110 mg vs. VKA 16.4%/year vs. 18.7%/year HR 0.87, 0.78–0.97 Warfarin vs. no therapy POSITIVE NCB Warfarin vs. aspirin POSITIVE NCB Rivaroxaban 15 mg vs. VKA Neutral Dabigatran 110 mg vs. VKA Positive AF, atrial fibrillation; GI, gastrointestinal; HR, hazard ratio; ICH, intracranial haemorrhage; NCB, net clinical benefit; NOAC, non-vitamin K antagonist oral anticoagulant; OAC, oral anticoagulant; SE, systemic embolism; VKA, vitamin K antagonist. Calculated using three different methods (Taiwanese cohort) or assessed as a composite outcome of all-cause death, major bleeding, and stroke/systemic embolism (French cohort). Major findings from ‘real-world’ atrial fibrillation cohorts of very old patients with atrial fibrillation 4.4% in the ‘VKA era’ (1996–2011) 8.6% in the ‘NOAC era’ (2012–15) 3.9% in the ‘VKA era’ 16.1% in the ‘NOAC era’ VKAs vs. no therapy Stroke risk reduction with no significant difference in major bleeding/ICH Ischaemic stroke 3.83%/year vs. 7.01%/year HR 0.61, 0.40–0.94 ICH HR 1.46, 0.57–3.71 NOACs vs. warfarin—effectiveness Similar or better effectiveness of NOACs in stroke risk reduction as compared to VKAs Ischaemic stroke 4.07%/year vs. 4.59%/year HR 1.16, 0.61–2.22 Stroke/SE Rivaroxaban 15 mg vs. VKA 3.8%/year vs. 3.3%/year HR 1.14, 0.93–1.40 Dabigatran 110 mg vs. VKA 2.8% vs. 3.6% HR 0.77, 0.60–0.99 NOACs vs. warfarin—safety Major bleeding Similar safety regarding major bleeding 6.14%/year vs. 6.75%/year HR 0.95, 0.63–1.44 Rivaroxaban 15 mg vs. VKA 3.7%/year vs. 4.0%/year HR 0.91, 0.74–1.11 Dabigatran 110 mg vs. VKA 3.3% vs. 4.0% HR 0.81, 0.64–1.03 NOACs vs. warfarin—safety ICH Significantly greater safety regarding ICH 0.42%/year vs. 1.63%/year HR 0.32, 0.10–0.97 Rivaroxaban 15 mg vs. VKA 0.8%/year vs. 1.1%/year HR 0.80, 0.54–1.20 Dabigatran 110 mg vs. VKA 0.5%/year vs. 1.2%/year HR 0.38, 0.22–0.66 NOACs vs. warfarin—safety Gastrointestinal bleeding Similar or greater risk for GI major bleeding with NOACs compared to VKAs Rivaroxaban 15 mg vs. VKA 2.5%/year vs. 2.0%/year HR 1.27, 0.98–1.65 Dabigatran 110 mg vs. VKA 3.2%/year vs. 1.7%/year HR 1.87, 1.40–2.51 All-cause mortality Similar or lower all-cause mortality with NOACs in comparison to VKAs Rivaroxaban 15 mg vs. VKA 17.5%/year vs. 19.1%/year HR 0.91, 0.83–1.00 Dabigatran 110 mg vs. VKA 16.4%/year vs. 18.7%/year HR 0.87, 0.78–0.97 Warfarin vs. no therapy POSITIVE NCB Warfarin vs. aspirin POSITIVE NCB Rivaroxaban 15 mg vs. VKA Neutral Dabigatran 110 mg vs. VKA Positive 4.4% in the ‘VKA era’ (1996–2011) 8.6% in the ‘NOAC era’ (2012–15) 3.9% in the ‘VKA era’ 16.1% in the ‘NOAC era’ VKAs vs. no therapy Stroke risk reduction with no significant difference in major bleeding/ICH Ischaemic stroke 3.83%/year vs. 7.01%/year HR 0.61, 0.40–0.94 ICH HR 1.46, 0.57–3.71 NOACs vs. warfarin—effectiveness Similar or better effectiveness of NOACs in stroke risk reduction as compared to VKAs Ischaemic stroke 4.07%/year vs. 4.59%/year HR 1.16, 0.61–2.22 Stroke/SE Rivaroxaban 15 mg vs. VKA 3.8%/year vs. 3.3%/year HR 1.14, 0.93–1.40 Dabigatran 110 mg vs. VKA 2.8% vs. 3.6% HR 0.77, 0.60–0.99 NOACs vs. warfarin—safety Major bleeding Similar safety regarding major bleeding 6.14%/year vs. 6.75%/year HR 0.95, 0.63–1.44 Rivaroxaban 15 mg vs. VKA 3.7%/year vs. 4.0%/year HR 0.91, 0.74–1.11 Dabigatran 110 mg vs. VKA 3.3% vs. 4.0% HR 0.81, 0.64–1.03 NOACs vs. warfarin—safety ICH Significantly greater safety regarding ICH 0.42%/year vs. 1.63%/year HR 0.32, 0.10–0.97 Rivaroxaban 15 mg vs. VKA 0.8%/year vs. 1.1%/year HR 0.80, 0.54–1.20 Dabigatran 110 mg vs. VKA 0.5%/year vs. 1.2%/year HR 0.38, 0.22–0.66 NOACs vs. warfarin—safety Gastrointestinal bleeding Similar or greater risk for GI major bleeding with NOACs compared to VKAs Rivaroxaban 15 mg vs. VKA 2.5%/year vs. 2.0%/year HR 1.27, 0.98–1.65 Dabigatran 110 mg vs. VKA 3.2%/year vs. 1.7%/year HR 1.87, 1.40–2.51 All-cause mortality Similar or lower all-cause mortality with NOACs in comparison to VKAs Rivaroxaban 15 mg vs. VKA 17.5%/year vs. 19.1%/year HR 0.91, 0.83–1.00 Dabigatran 110 mg vs. VKA 16.4%/year vs. 18.7%/year HR 0.87, 0.78–0.97 Warfarin vs. no therapy POSITIVE NCB Warfarin vs. aspirin POSITIVE NCB Rivaroxaban 15 mg vs. VKA Neutral Dabigatran 110 mg vs. VKA Positive AF, atrial fibrillation; GI, gastrointestinal; HR, hazard ratio; ICH, intracranial haemorrhage; NCB, net clinical benefit; NOAC, non-vitamin K antagonist oral anticoagulant; OAC, oral anticoagulant; SE, systemic embolism; VKA, vitamin K antagonist. Calculated using three different methods (Taiwanese cohort) or assessed as a composite outcome of all-cause death, major bleeding, and stroke/systemic embolism (French cohort). In this issue of Europace, Fauchier et al.11 reported a retrospective analysis of the effectiveness and safety of reduced dose rivaroxaban (15 mg od) and lower dose dabigatran (110 mg bid) vs. VKAs (mostly flunidione) in the French nationwide cohort of very old AF patients (≥85 years, total n = 51 038) newly prescribed OAC, broadly confirming previously published reports (Table 1). The study also compared rivaroxaban and dabigatran, but such direct comparisons have numerous limitations and are difficult to interpret.12 The French nationwide cohort study included large number of very old AF patients and used a high-dimensional propensity score matching in order to diminish possible selection bias.12 Unfortunately, details on renal function were not available thus precluding differentiation between appropriate and off-label rivaroxaban dose reduction, whilst dabigatran dosage was appropriate for the participants’ age (see Figure 1, lower panel). Since renal function declines with aging,13 the use of rivaroxaban 15 mg od was likely appropriate in many participants but those with preserved renal function would be underdosed. Appropriate dose selection (Figure 1, lower panel) is crucial for optimal effects of NOACs.6 Evidence from ‘real-world’ datasets suggest that inappropriate use of lower or reduced NOAC doses in clinical practice may be associated with increased rates of thrombo-embolic events with no change in major bleeding (especially with apixaban) and diminished safety with no efficacy benefit in patients with severe renal failure.6 Another important determinant of optimal OAC use is adherence (and persistence) to treatment. Whereas adherence to VKAs is indirectly reflected by a good time in therapeutic range of ≥70%, various tools for optimizing the assessment and maintenance of adherence to NOACs in clinical practice are currently being studied.14 Of note, the information about adherence to OAC or drug discontinuation rates was not provided by Fauchier et al., but in a recent report from a large German dataset (n = 51 606) of AF patients initiating OAC, age ≥85 years was associated with a significantly lower likelihood of OAC discontinuation.15 Overall, the study of Fauchier et al. strengthens the contemporary evidence of (N)OACs effectiveness and safety in AF-related stroke prevention by providing reassuring observations from a large cohort of very old patients with AF. Such patients have a higher absolute risk of stroke and other adverse outcomes compared with their younger counterparts with AF, and the higher the risk the greater absolute risk reduction with (N)OACs. Age is a strong independent risk factor for stroke and very old AF patients should be considered for OAC, preferably NOACs, with concomitant bleeding risk assessment. Increased bleeding risk itself is not a reason to deny OAC to the elderly with AF, but modifiable bleeding risk factors should be identified and corrected, and high-risk patients should be scheduled for a more intense clinical follow-up with bleeding risk re-assessment at every clinical visit. Conflict of interest: none declared. The opinions expressed in this article are not necessarily those of the Editors of Europace or of the European Society of Cardiology.