Aging With Atrial Fibrillation: The Dilemma of Anticoagulation

Older adults with multiple comorbidities are at a higher risk for thromboembolic events compared with the general population. Currently, the major predictor of cardiovascular events in older individuals is atrial fibrillation (AF).¹ AF is an arrhythmia characterized by the rapid and irregular beating of the atria, leading to heart palpitations, chest pain, and lightheadedness.² According to the Framingham study, the 10-year mortality rate for older adults with AF is 61.5% in men and 57.6% in women. The mortality rate attributable to AF is three times higher in adults older than 75 years than those younger than 65 years.³ AF is more likely to occur in older adults, as it is present in roughly 2% of patients younger than 65 years and 9% of patients older than 65 years. Additionally, the annual stroke risk due to AF is 23.5% for patients between ages 80 and 89 years, significantly higher than the 1.5% risk for patients between ages 50 and 59 years.⁴

In 2016, among the approximately 46 million people in the United States older than 65 years, 4.14 million (9%) were estimated by the Centers for Disease Control and Prevention (CDC) to have AF.^2,5 This population is projected to increase to 98 million by the year 2060, reflecting a total population shared increase from 15% to 24% and implying that the number of older adults with AF and associated health care costs will also likely increase.⁵ The Cardiovascular Health Study and the Framingham Heart Study analyzed the costs associated with newly diagnosed AF in 513 participants aged 65 years or older; they found that the mean annual health care costs increased from $5158 to $25,675 with the diagnosis of AF, with values adjusted for 2009 dollars.⁶ This increase in annual costs per patient also accounted for other costly conditions that are associated with AF, including stroke and heart failure. A separate retrospective, observational cohort study⁷ found that the total incremental cost of AF alone was $8705. The total additional cost imposed by AF can then be calculated by multiplying this incremental cost by the estimated 4.14 million people older than 65 years with AF, resulting in roughly $36 billion in total additional cost, much higher than the $6.1 billion estimated by the CDC.²

Due to the high morbidity, mortality rates, and costs associated with this condition, multiple management options have been developed for AF. These include surgery to fix the underlying structural heart disease, medications to control the heart rate and rhythm, and anticoagulation therapy to reduce stroke risk.⁸ Studies have established that anticoagulation is the optimal therapy for stroke prevention in patients with AF.¹ The event rate for strokes and thromboembolisms increases with age and is significantly higher in adults older than 75 years.⁹ The primary disadvantage of anticoagulation therapy is the increased bleeding risk, which is more pronounced in older adults with multiple comorbidities.¹⁰ However, in a study involving 229 older adults with a mean age of 73 years who had nonrheumatic AF, there was no change in the incidence of major bleeding, despite an increase in the number of patients receiving anticoagulation therapy from 14% to 34%. This study also found lower mortality rates among older adults on oral anticoagulation therapy (19%) when compared to those on antiplatelet therapy (32%) and those without any antithrombotic therapy (67%).¹¹

Despite this data, there are still reports of general tendencies to underuse oral anticoagulants in older adults with AF.¹² Therefore, the notion that older adults have higher bleeding risks when placed on anticoagulation therapy should be further investigated in order to provide optimal care for the older population. In this article, we discuss the different types of oral anticoagulants, their indications and contraindications, and specific considerations for anticoagulation in older patients with AF.

The four main classes of anticoagulants include vitamin K antagonists (VKAs), nonvitamin K oral anticoagulants (NOACs), heparins, and synthetic pentasaccharides. These medications are generally indicated for AF, acute coronary syndrome, heart failure, stroke, deep vein thrombosis, and pulmonary embolism. They share similar contraindications, such as patients with large esophageal varices or other clinically significant bleeding conditions, 72 hours post-surgery, neurosurgery within 3 months, trauma, and thrombocytopenic state.¹³ The most concerning complication of anticoagulation therapy for both patients and providers is bleeding. Additionally, the risk of bleeding with each specific anticoagulant is determined by multiple factors. The availability of corresponding reversal agents is therefore very important, but only some anticoagulants have reversal agents.

Vitamin K Antagonists

VKAs are oral anticoagulants that are structurally similar to vitamin K and act as competitive inhibitors of vitamin K-epoxide reductase. This competitive inhibition blocks the recycling of vitamin K, a component required for activating coagulation factors II, VII, IX, and X, and proteins C and S.¹⁴ Warfarin is the most commonly used VKA worldwide. The dosing of warfarin is guided by a goal of international normalized ratio (INR) between 2 and 3. Its main limitation is its narrow therapeutic range. Time spent below or above the INR therapeutic range significantly increases the risk for thromboembolic or bleeding events, respectively. This is especially important in adults aged 60 years and older, who generally require a lower dosage to achieve a therapeutic INR due to changes in warfarin metabolism. The reversal agents for warfarin are phytonadione (vitamin K₁), fresh frozen plasma, and prothrombin complex concentrate.^15,16

NonVitamin K Oral Anticoagulants

NOACs include both direct thrombin (factor IIa) and factor Xa inhibitors. They provide oral anticoagulant options that are more pharmacokinetically and pharmacodynamically stable than VKAs, with a faster onset/offset of action and fewer drug interactions. They do not require INR monitoring and thus have become the preferred medication among patients without compromising medication adherence. They have also been shown through clinical studies to be as efficacious and safe as VKAs at preventing stroke in AF as well as treating and preventing venous thromboembolism. Despite these advantages, certain NOACs cannot be used in patients with impaired kidney function due to the increased risk of bleeding (Table 1).¹⁷ 

Direct thrombin inhibitors (DTIs) block thrombin, the final enzyme of the clotting cascade that produces fibrin. There are three types of DTIs and they each interact differently with thrombin. Divalent DTIs—which include hirudin, bivalirudin, lepirudin, and desirudin—bind to active site and exosite 1. Univalent DTIs—which include argatroban, inogatran, melagatran, and dabigatran—bind only to the active site. The third type of DTI are the allosteric inhibitors.¹⁸ Previously, the largest disadvantage for these drugs was the lack of a reversal agent. In 2016, the Food and Drug Administration (FDA) approved the first-ever NOAC reversal agent, idarucizumab, for the reversal of the anticoagulant effects of dabigatran.¹⁹

Direct factor Xa inhibitors block factor Xa, which normally cleaves prothrombin to yield an active thrombin. These drugs include rivaroxaban, apixaban, betrixaban, and edoxaban. In 2018, the FDA approved a reversal agent, specifically for rivaroxaban and apixaban, called andexanet alfa.²⁰ The ANNEXA-4 clinical trial found andexanet alfa to be highly effective in reversing rivaroxaban and apixaban, reducing both of their bioavailabilities by 92% within 2 hours of administration.²¹ With a cost of up to $50,000, this agent currently has limited availability.²² 

Heparins

The two types of heparins are the parenteral unfractionated heparin (UFH) and the subcutaneous low-molecular-weight heparins (LMWH). Heparins potentiate the ability of antithrombin III to inactivate thrombin, as well as coagulation factors IIa, IXa, Xa, XIa, and plasmin, and ultimately prevent conversion of fibrinogen to fibrin. Some of the available LMWHs include enoxaparin, dalteparin, and certoparin. UFH inactivates factors Xa and IIa at a ratio of 1:1, whereas LMWHs have a higher impact on factor Xa than on factor IIa, with ratios ranging between 2:1 and 4:1.²³

A meta-analysis of 13 randomized controlled trials compared UFH and LMWHs and found LMWHs to be more efficacious at preventing recurrent venous thromboembolisms and less prone to major bleeding. LMWHs can be administered in the outpatient setting while UFH requires inpatient hospitalization for administration. Due to LMWHs’ generally smaller size, they do not interact with platelet-factor-4 and platelets as closely as UFH does, and this significantly decreases the risk for heparin-induced thrombocytopenia, a feared complication of heparins.²⁴ Because of these multiple advantages, LMWHs are typically preferred over UFH. However, LMWHs rely more heavily on renal excretion and are thus less suitable for patients with renal dysfunction.²³

Synthetic Pentasaccharides

Synthetic pentasaccharides include fondaparinux and idraparinux, which are parenteral indirect factor Xa inhibitors. These drugs bind to antithrombin and accelerate its inhibition of factor Xa. Their main advantage over LMWH and UFH is a significantly lower risk of heparin-induced thrombocytopenia, but they are not used in patients with renal dysfunction due to their renal excretion.¹⁴ Idraparinux has an elimination half-life that is 5 to 6 times longer than fondaparinux and thus provides the benefit of requiring only one injection per week rather than daily injection.²⁵

 AF is generally categorized as paroxysmal, persistent, or permanent. Paroxysmal AF is defined as variable intermittent AF that either self-terminates or terminates with intervention within 7 days of onset. Persistent AF fails to terminate within 7 days and often requires restoration of sinus rhythm with pharmacologic or electrical cardioversion. Permanent AF is persistent and requires a joint decision by the patient and the provider to no longer pursue a rhythm control strategy.²⁶ 

Patients with AF were previously categorized as having either valvular or nonvalvular AF, but the term “nonvalvular” is no longer considered appropriate. Instead, patients are differentiated based on whether or not they have a history of moderate to severe mitral stenosis or a mechanical heart valve. The 2019 Focused Update on Atrial Fibrillation guideline suggests that all patients with AF, irrespective of their category, should be anticoagulated based on their risk of thromboembolism, and initiation of therapy should be individualized based on a patient’s absolute and relative risks of stroke and bleeding.²⁷

According to the updated AF guidelines, the stroke risk for patients with AF should be determined using the CHA₂DS₂-VASc score, except for those with moderate to severe mitral stenosis or a mechanical heart valve (Table 2). This scoring system includes the stroke risk modifiers of congestive heart failure or left ventricular systolic dysfunction, hypertension, age greater than or equal to 75 years, diabetes mellitus, previous stroke or transient ischemic attack or thromboembolism, vascular disease, age between 65 to 74 years, and sex category. Patients with AF and a CHA₂DS₂-VASc score of 2 or greater in men and 3 or greater in women should be anticoagulated with oral anticoagulants, according to the scoring system. It is only reasonable to omit anticoagulation therapy in patients without moderate to severe mitral stenosis or a mechanical heart valve and a CHA₂DS₂-VASc score of 0 in men or 1 in women.²⁷

The options for oral anticoagulation include warfarin, dabigatran, rivaroxaban, apixaban, and edoxaban, which was approved by the FDA in 2015 after the ENGAGE AF-TIMI 48 trial demonstrated its noninferiority to warfarin for stroke prophylaxis in patients with AF.²⁸ The class I recommendations include that NOACs are now recommended over warfarin for patients with AF, except for those with moderate to severe mitral stenosis or a mechanical heart valve. This is because the NOAC trials revealed that these agents are either noninferior or superior to warfarin for preventing stroke and have a lower risk of serious bleeding.²⁷ These agents do not require weekly determination of the INR, whereas warfarin does. All patients with moderate to severe mitral stenosis or a mechanical heart valve should still be anticoagulated, preferably with warfarin.

Anticoagulation in older adults, especially those in long-term care (LTC) settings, requires a careful balance between stroke prevention and hemorrhage. Although systemic anticoagulation is recommended when the CHA₂DS₂-VASc score is greater than 2, older populations warrant specific consideration prior to initiation of anticoagulation therapy. Consequently, many clinicians are hesitant to initiate anticoagulation; studies have shown that only 56% of patients older than 65 years with AF receive anticoagulation.²⁹ In the LTC setting, 7.5% to 17% of patients have an established diagnosis of AF.³⁰ The prevalence of AF in older adults in LTC settings highlights the importance of weighing the risks and benefits prior to initiation of anticoagulation. 

The risk of falls and subsequent bleeding are the most common causes for not using anticoagulation in older adults.³¹ It is prudent to evaluate fall risk because of the pervasiveness of falls in older adults, as an estimated 30% to 40% of independent individuals older than 65 years and up to 50% of those living in LTC facilities experience at least one fall per year.³² A major factor contributing to falls is the prescription of medications such as antipsychotics, antidepressants, hypnotics, diuretics, and anticholinergics. Additionally, polypharmacy significantly increases the risk of falls and hemorrhage, as up to 36% of patients in LTC facilities take nine or more regular medications.³³ Falls are the leading cause of fatal and nonfatal injuries in older patients, and in 2014, approximately 27,000 older patients died due to complications from mechanical falls.^34,35 Thus, there is an argument to be made for starting systemic anticoagulation in older adults older than 65 years, as falls are associated with an increased risk of bleeding events. The average risk of stroke per year in older adults with AF is about 5%.³⁶ Despite this, a given patient would have to fall approximately 295 times for the risks of anticoagulation to outweigh the benefits of stroke prevention.^10,37

Another consideration when evaluating fall risk is baseline cognitive functional ability to perform activities of daily living (ADLs). Inability to perform ADLs and poor cognitive function contribute to an increased risk of falls. As a result, a holistic approach is necessary to evaluate social support for high-risk patients with AF who require anticoagulation for stroke prevention in order to reduce their risk of falls. Although fall risk and a history of falls are not absolute contraindications to systemic anticoagulation, it appears that the benefits of systemic anticoagulation for stroke prevention outweigh the potential risks associated with falling in older adults. Ultimately, it is important to utilize a shared-decision making approach when deciding to start an older adult with a history of falls on systemic anticoagulation and to implement strategies to prevent future falls. Although LTC settings provide more surveillance, monitoring, and support for older adults to help mitigate fall risks, the shared-decision making model is crucial. While the risk of falls and subsequent potential hemorrhage are the most significant dangers of starting anticoagulation in older adults, there are several other risks to consider. 

Another factor for anticoagulation in older adults is anemia. Studies have shown that 48% to 63% of older individuals in LTC experience anemia, which directly contributes to an increased risk of bleeding. Normally, erythrocytes force platelets centrifugally toward the vessel wall, which facilitates platelet adhesion and clot formation when vascular integrity is disrupted.³⁸ In patients with anemia, the decreased number of erythrocytes impairs this interaction, and thus platelet adhesion is decreased. Another major risk factor is the high prevalence of renal disease in older adults. The prevalence of chronic kidney disease (CKD) in the US adult population is approximately 11%; however, the prevalence in older adults is much higher, estimated to be 39.4%.³⁹ Furthermore, 45.7% of patients living in LTC settings have nondialysis dependent CKD.⁴⁰ Stage 3 CKD is associated with twice the rate of major bleeding during anticoagulation with warfarin due to reduced clearance of the drug. Additionally, older adults have an increased sensitivity to warfarin as they are more likely to have fluctuations in their INR due to polypharmacy regimens, poor diet, disease comorbidities, and compliance issues.^10,37 With the increased lability of the INR, older adults are less likely to be within the therapeutic range at any given point when compared to the general population and thus are at an increased risk of major bleeding events. 

The CHA₂DS₂-VASc score should be utilized for patients with AF to determine the necessity of anticoagulation therapy, except for those with moderate to severe mitral stenosis or a mechanical heart valve. Additionally, patients with AF and moderate to severe mitral stenosis or a mechanical heart valve should be anticoagulated. Studies have shown that NOACs are at least as effective as warfarin at preventing stroke and systemic embolism and have a lower risk of major bleeding. As such, NOACs are now recommended over warfarin in patients with AF who do not have moderate to severe mitral stenosis or a mechanical heart valve. 

Although older adults with AF have a significantly higher risk than younger adults of debilitating complications such as stroke and thromboembolism, only half of patients older than 65 years receive anticoagulation therapy. In LTC patients, it is critical to perform a careful and meticulous evaluation of the risks and benefits through shared decision-making to achieve an appropriate balance between stroke prevention and bleeding risk. Ultimately, the benefits of stroke prevention may outweigh these risks because the event rates of stroke, thromboembolism, and death are greatly increased in older patients in LTC settings.

Affiliations:
¹ Lake Erie College of Osteopathic Medicine,
Erie, PA
² College of Medicine, Pennsylvania State
University, Hershey, PA
³ Milton S Hershey Medical Center,
Pennsylvania State University, Hershey, PA

Disclosures: The authors report no relevant financial relationships.

Address correspondence to:
Zainab Shahid
Lake Erie College of Osteopathic Medicine
1858 W. Grandview Blvd
Erie, PA 16509
Phone: (718) 578-0296
Fax: (717) 531-7726
Email: zshahid44251@med.lecom.edu