Journal of Thrombosis and Thrombolysis

, Volume 27, Issue 4, pp 461–464

Warfarin in vulnerable older adults with atrial fibrillation


  • S. Michael Gharacholou
    • Division of CardiologyDuke University Medical Center and Duke Clinical Research Institute
    • Division of GeriatricsDuke University Medical Center and Duke Clinical Research Institute
    • Division of CardiologyDuke University Medical Center and Duke Clinical Research Institute
    • Duke Clinical Research Institute

DOI: 10.1007/s11239-009-0344-0

Cite this article as:
Gharacholou, S.M. & Becker, R.C. J Thromb Thrombolysis (2009) 27: 461. doi:10.1007/s11239-009-0344-0

Current indications for anticoagulant therapy in the primary or secondary prevention of thromboembolism span a vast array of clinical syndromes in cardiovascular medicine, from nonvalvular atrial fibrillation (AF) to prosthetic mechanical valves, and the prophylaxis of arterial or venous thrombosis [1, 2]. The complexities of vitamin K antagonist (VKA) administration among patients with AF is heightened substantially in older adults who, in addition to comborbid illness, pose a risk for both thrombotic and hemorrhagic events-either of which can be life-altering or life-threatening.

VKAs, of which warfarin is the most widely used and investigated coumarin derivative, exerts its anticoagulant effects though modulating γ-carboxylation of glutamic acid (GLA) residues of the vitamin K-dependent coagulation factors II, VII, IX, and X, resulting in the production of coagulation factors with reduced coagulant capacity [3]. The estimated disease prevalence of AF is 9% among octogenarians, and given global aging, the total AF burden of disease is expected to double by 2050 [4]. Complexities in decision making around the use of VKAs are magnified among the population of vulnerable older adults with functional decline, nutritional deficiencies, polypharmacy, frailty, falls, and cognitive dysfunction—clinical and health-related characteristics that collectively define the older adult with prototypical geriatric syndromes [57]. Vulnerability has been previously operationalized and defined as “community-dwelling seniors who have four times the risk for functional decline or death over the next 2 years compared with those not identified as vulnerable” [8]. This phenotype has previously been captured, employing either self or proxy reporting on aspects of functional status, and collectively represent conditions frequently encountered in the practice of geriatrics [9]. Accordingly, the direct application of evidence from existing guidelines and clinical studies on the benefits of anticoagulant therapy for conditions, like AF, that are known to increase the risk of thrombosis-related clinical events must be balanced carefully by considering both the number needed to harm (NNH) and the “values and preferences” of care for an individual patient. The absolute benefit of VKA over antiplatelet therapy among the very elderly with geriatric syndromes continues to be an area of uncertainty, despite being encountered frequently in clinical practice.

Building upon the overarching theme of hemostasis and thrombosis in older adults [10], herein we consider the landscape of VKA treatment among vulnerable older adults—offering a paradigm of thrombotic risk, bleeding propensity, and individualized decision making to include cessation of treatment when the likelihood of harm outweighs the potential benefit.

Navigating the benefits and risks of treatment: a continuum of care

Despite a reduction of stroke risk among patients with nonvalvular AF receiving aspirin (approximately 20% risk reduction), and recent observations proposing a potential benefit to dual antiplatelet-directed therapy among patient’s deemed less desirable (i.e., bleeding risk, physician, or patient preference) for VKA treatment [11], VKA remains the most effective drug for primary stroke prevention among older adults with AF, reducing the risk of stroke or transient ischemic attack by one-third when compared with antiplatelet therapy alone [12]. Randomized controlled trials have demonstrated the ability of VKA to reduce the annual stroke rate from 4.5% to 1.5% [13]. The CHADS2 score, a widely employed risk prediction tool, provides a valuable reference point for clinicians trying to gauge the likelihood of stroke without anticoagulation, increasing by a factor of 1.5 for each point in the 6-point scale [14]. Clinical practice guidelines from the American Geriatrics Society [15] have reflected previous recommendations for anticoagulation to minimize the risk of thromboembolism associated with AF issued by the American College of Chest Physicians [16]. Given the high rates of morbidity, disability, cognitive impairment, and mortality associated with cardioembolic, ischemic stroke, older adults with AF, whether paroxysmal, chronic or permanent in nature, should be considered carefully for VKA anticoagulation.

A comprehensive evaluation also must determine the strategy, safety, feasibility for stringent anticoagulation monitoring, bleeding risk, and patient-centered preferences for anticoagulation therapy. Despite the overall disease prevalence, octogenarians have limited representation in randomized trials of anticoagulation for AF. Results from single center studies have demonstrated that, along with increasing CHADS2 score comes a greater risk of major bleeding, typically pronounced in those above target international normalized ratios (INRs), and high rates of VKA discontinuation due to both overt bleeding and perceived bleeding risk [17, 18]. A clustering of major bleeding events occurs within the first 90 days of VKA initiation. In addition, major bleeding occurs nearly three times as often in patients ≥80 years of age [17]. Among patients with a CHADS2 score >4, those ≥85 years of age experience a NNH of 10 for major bleeding [18]. Similarly, when clinicians recommend antiplatelet therapy for either the primary or secondary prevention of vascular events in older adults, there must be an awareness of concomitant bleeding risk when drug exposure occurs over an extended period of time [19]. The exclusion of older patients from clinical trials who are managed within long-term care facilities; inconsistent inclusion of head trauma when considering rates of intracranial hemorrhage; and, lack of a standard assessment tool for determining fall risk present major challenges for selecting older adults in whom the benefit of treatment may not outweigh the risk. Emphasis on assessment of bleeding risk should preferably incorporate the use of bleeding risk models or classification tools, such as the Outpatient Bleeding Risk Index or the HEMORR2HAGES score [20], which can aid in the overall decision-making process. These “processes of care” may further support individualized decision making among clinicians and anticoagulation clinics, which often employ a construct of “equipoise” for cessation of anticoagulant therapy when the anticipated risk either equals or exceeds the benefit.

Insight may be gained through comprehensive geriatric assessment (CGA) [21], a process wherein geriatricians, geriatric nurse specialists, pharmacists with expertise in geriatrics, and dietitians familiar with geriatric nutritional health, provide assessment and recommendations for older adults with geriatric syndromes. Home visits to evaluate patients and their surroundings for potential “red flags” associated with heightened fall risk and a targeted physical therapy evaluation to assess gait, strength, and balance for planning and implementing individualized goals may be particularly beneficial. A thorough evaluation can further clarify concerns regarding specific geriatric conditions and circumstances (i.e., falls, cognitive dysfunction, and polypharmacy) [22], which are often cited as reasons not to prescribe, or to discontinue VKA-based anticoagulation. A comprehensive approach to geriatric syndromes may, in fundamental terms, provide the information required to ascertain the balance of benefit and risk and to craft a generalizable approach for clinical decisions. The role of a geriatric-centered multidisciplinary strategy to guide antithrombotic therapy in AF, and its effect on patient outcomes, has not been evaluated previously; however, this approach is conceptually attractive in complex cases (Fig. 1). A highly structured and step-wise approach to the evaluation and management of older adults can (and should) be applied to both standard VKA-based treatment as well as emerging oral anticoagulants. Indeed, it is counterintuitive to use incident events, such as intracranial hemorrhage or life-threatening bleeding as the sole criterion for cessation of anticoagulants.
Fig. 1

Patient-centered model of factors for characterizing risk, benefit, and optimized selection of anticoagulant therapy candidacy among older adults with atrial fibrillation. Asterisk represents dose reduction in patients with heart failure and congestive hepatopathy

Despite being over 50 years since their first clinical application and a proven track record of efficacy for a variety of thrombotic disorders, VKAs have complex pharmacokinetic and pharmacodynamic properties that demand high-level dose-adjusted monitoring and expertise. Genetic polymorphisms of CYP2C9 and VKORC1 causing substantial variability in drug dosing [23], and pharmacogenetically guided approaches to dose selection may be useful among patients at high risk for thrombotic and hemorrhagic events [24].

Several anticoagulants with target selectivity, limited variability, minimal (or less compared to VKA) pharmacokinetic and pharmacodynamic interactions, and fixed dosing potential are in advanced stages of development [25]. In the ROCKET-AF trial, rivaroxaban, an oral direct factor Xa inhibitor given at a dose of 20 mg once daily (or 15 mg daily in subjects with a reduced creatinine clearance), is being compared against warfarin (active control) for stroke prevention in 14,000 patients [26]. In ARISTOTLE, the oral direct factor Xa inhibitor apixiban given at a dose of 5 mg twice daily is being compared to warfarin in 15,000 patients for the primary outcome of stroke or systemic embolism in patients with AF [27]. Given that both trials have included age ≥75 years as an enrichment criteria, the anticipated number of older adults being treated with oral anticoagulants will provide a much needed data set for understanding AF as a disease of aging and optimized, patient-specific management [26, 27]. A note of caution is warranted—if one or more contemporary anticoagulants successfully achieve FDA approval, careful selection and management of older adults will remain a cornerstone of treatment. While routine coagulation monitoring will not be required, intermittent “in-person” patient evaluation, monitoring of renal function, and continued patient education should be recommended.

Copyright information

© Springer Science+Business Media, LLC 2009