55: Considering Anticoagulation in Older Adults

Persons 65 years or older comprise approximately 13% of the U.S. population, yet they are prescribed the greatest proportion of medication; 90% are prescribed at least 1 prescription drug and 65% 3 or more prescription drugs in the last month. They also represent the fastest growing segment of prescription drug users in the United States. Thus, with an increasing number of patients surviving to older age and consuming larger amounts of drugs, it is necessary for clinicians to understand the risk, benefits, and consequences of drug therapy in older patients. This is especially true for anticoagulants, a class of drugs essential for the optimal management of many thromboembolic and vascular disorders that are highly prevalent among older patients. Anticoagulants are unique compared to most pharmacologic agents because even small deviations from “therapeutic levels” place patients at risk for life-threatening complications. While older patients with multimorbidity are particularly susceptible to thrombosis, they also have higher risks of bleeding than the general population. This chapter briefly reviews current anticoagulant therapy and focuses on the newer agents and recommendations for their use in older patients.

Current anticoagulants that are available for use within the United States include unfractionated heparin (UFH), vitamin K antagonists (VKAs), low-molecular weight heparins (LMWHs), indirect selective factor Xa inhibitor, and the direct thrombin (parental and oral) and factor Xa inhibitors. Tables 55–1 and 55–2 summarize the specific pharmacologic characteristics of these agents.

Oral Vitamin K Antagonists

Concerns about the use of anticoagulants in older patients arise from their increased risk for anticoagulant-related bleeding. The major determinants of oral VKA-induced bleeding are the intensity of the anticoagulant effect, as measured by the international normalized ratio (INR), patient characteristics, concomitant use of drugs that interfere with hemostasis or vitamin K metabolism, and the length of therapy. Of these, the INR is the most important risk factor, and this is especially true for the intracranial hemorrhage (ICH), the most feared site of major bleeding. The risk of ICH increases 7-fold with increasing INR levels above 4.0. Patient characteristics, including age and specific comorbid conditions (ischemic stroke diabetes, renal insufficiency, malignancy, hypertension, liver disease or alcoholism), also are associated with increased risk of major bleeding. In general, older patients have approximately a 2-fold increase in major bleeding compared to their younger counterparts. Decision making around the use of anticoagulants is complex because the risk factors that are associated with anticoagulant-related bleeding are similar to those associated with increased risk of thrombosis. The use of anticoagulants medications in older patients is an area where applying the principles of shared decision making is critical. The choice of whether to prescribe anticoagulants and which specific medication to use should be individualized, taking into account not only evidence-based medicine, but also patients’ goals and preferences to ensure compliance.

More recently, the genetic polymorphisms of the cytochrome P450 CYP2C9 and the vitamin K epoxide reductase complex subunit 1 (VKORC1) have been found to affect warfarin metabolism and vitamin K reduction. Patients with the CYP2C9*2, CYP2C9*3, and VKORC1 A variants require lower maintenance doses of warfarin and are at increased risk of over-anticoagulation and major bleeding. Thus, on average, dosage reductions of approximately 19% to 33% are needed to avoid over-anticoagulation. The International Warfarin Pharmacogenetics Consortium found that an algorithm based on clinical plus pharmacogenetic data performed better compared to either pure clinical algorithm or fixed dose approach in predicting the appropriate warfarin dose. Subsequently, the Food and Drug Administration revised the labeling for warfarin in 2010 to reflect that CYP2C9 and VKORC1 genotyping could assist in warfarin dosing. Yet, the clinical utility of pharmacogenetic-based warfarin dosing in older patients is not clear. Schwartz et al noted in a cohort of patients ≥65 years (mean age: 81 years) that included nursing home and senior care community residents on warfarin with stable therapeutic INRs that the addition of genotype information helped to explain a greater proportion of the INR variability when compared to those without genotype information (50% vs. 12%; P <0.0001). However, when comparing estimated warfarin doses to actual warfarin doses in patients requiring <2 mg per day of warfarin, dosing was overestimated despite the use of pharmacogenetic information; that is, the addition of genotype information did not improve the dosing management. Because earlier studies observed increasing age is associated with increasing response to the effects of warfarin as manifested by lower daily doses, the applicability of pharmacogenetic dosing algorithms to older patients requiring lower warfarin dosing is somewhat limited. Thus, the adage of “start low and go slow” still remains applicable to warfarin dosing in older patients.

Many drugs are known to interact with anticoagulants, and because the majority of older patients are prescribed more than 1 drug, there is ample opportunity for adverse drug reactions to occur in older patients. Drugs that potentiate the anticoagulant effect (increase the INR) increase the risk of bleeding. Other drugs increase hepatic metabolism resulting in decreasing the anticoagulant effect requiring increased dosage requirements (Table 55–3). When these drugs are discontinued, there can be an increase in INR and bleeding. Thus, additional monitoring with dosage adjustment is required when these drugs are either added to or removed from the medication profile of older patients on warfarin therapy.

Despite its efficacy in treatment and prophylaxis, warfarin has several limitations that make its use cumbersome. These include its slow onset of action, narrow therapeutic window, and the lack of predictability in anticoagulant effect by drug dose, many dietary and drug interactions, and the need for routine INR monitoring. Some of this burden may be lessened with less-frequent INR monitoring (up to every 12 weeks vs. every 4 weeks), which has been shown to be safe in patients with stable INRs. Older patients who are motivated and can demonstrate competency can self-manage and/or self-test. Best practices for ensuring safety include using a coordinated monitoring system with patient education, systematic INR testing, tracking and follow-up, and good communication. (For travel recommendations for older adults on warfarin, see Chapter 20, “The Aging Traveler.”)

Injectable Anticoagulants

LMWH and selective indirect anti-Xa inhibitor (fondaparinux) are also used in older patients. The 2 major concerns in older patients are renal impairment and lower body weight. Reduced renal clearance occurs with age and increases the susceptibility to major bleeding, as LMWHs are primarily renally eliminated. The risk of LMWH accumulation and bleeding is dependent on the severity of renal impairment, the dose (prophylactic or therapeutic) and type of LMWH. Among LMWHs, only enoxaparin has approved dose reduction in older patients with renal impairment. Advancing age and renal impairment also reduces clearance of fondaparinux. Reduced-dose fondaparinux appears to have good safety and efficacy in older patients with mild renal impairment, but this has not been validated in those with severe renal impairment. Renal function should not be solely assessed by the serum creatinine as this leads to underestimation of renal failure in older patients, and measurement of the glomerular filtration rate is preferred. Thus, it is prudent to test LMWH or fondaparinux anti-Xa levels in older patients with renal impairment or low body weight to avoid supratherapeutic doses.

New Oral Anticoagulants

For the first time since the introduction of warfarin in 1954, 2 new oral anticoagulants have been FDA approved. Dabigatran is FDA approved for the prevention of stroke and systemic embolism in nonvalvular atrial fibrillation. Rivaroxaban is FDA approved for (a) prophylaxis against venous thromboembolism (VTE) in patients undergoing knee or hip replacement surgery and (b) to reduce the risk of stroke and systemic embolism in patients with nonvalvular atrial fibrillation. Although clinical trials that led to the approval of dabigatran and rivaroxaban included older patients, those with renal and hepatic failure were systematically excluded from these trials. These new oral anticoagulants overcome several of the limitations of warfarin, including slow onset of action, narrow therapeutic window, drug and dietary interactions, and the need for routine laboratory monitoring. As a result of the increased use of these agents in the geriatric population, clinicians should be aware of the indications, pharmacology, methods for monitoring anticoagulant activity, and recommendations for management of bleeding with these new oral anticoagulants (see Table 55-2 and Tables 55–4, 55–5, and 55–6).

Dabigatran is a novel competitive direct thrombin inhibitor. Dabigatran has a rapid onset of action (peak plasma concentration approximately 1.5 hours following oral ingestion) obviating the need for bridging anticoagulation. The half-life of dabigatran with normal renal function is 12–14 hours, allowing for once or twice-daily dosing. Because the drug is 80% renally cleared, dose reductions (75 mg PO BID) are recommended for creatinine clearance (CrCl) of 15–30 mL/min. Dabigatran is contraindicated in patients with severe renal insufficiency (CrCl <15 mL/min). Consideration should be given to age-related changes in kidney function that lead to a 40% to 60% increase in area under the curve. This increased concentration can lead to greater exposure to the drug and potential bleeding complications. Age-related renal function changes may explain some of this age effect. In the RE-LY study, 2 doses of dabigatran, 110 mg and 150 mg twice daily, were compared to dose-adjusted warfarin in more than 18,000 patients with nonvalvular atrial fibrillation. The mean age in each group was 71 years and the mean weight 82 kg. Patients with CrCl <30 mL/min and clinically significant liver disease were excluded. The trial was therefore neither designed nor powered to detect the safety of dabigatran in older patients with low body weight. A significant proportion of patients in each group were on aspirin concomitantly. Both doses of dabigatran were proven to be as effective as warfarin for prevention of stroke or systemic embolism. However, dabigatran 150 mg twice daily was found to be superior to warfarin for stroke prevention (1.11% per year vs. 1.71% per year; p <0.001; RR: 0.65; 95% CI: 0.52–0.81). There was no difference in annual major bleeding rates on dabigatran compared to warfarin (3.32% per year vs. 3.57% per year; p = 0.32; RR: 0.93; CI: 0.81–1.07). However, a subset analysis by age groups revealed that among the 39% of patients older than 75 years old, bleeding was increased in the dabigatran 150 mg arm (HR: 1.18; 95% CI: 0.98–1.43). This effect occurred regardless of renal function. Although not Federal Drug Administration (FDA)-approved for the following indications, dabigatran was also found to be as effective as (a) enoxaparin in the prevention of VTE after total knee arthroplasty and total hip arthroplasty, and (b) warfarin in the prevention of recurrent VTE after acute symptomatic proximal deep venous thrombosis or pulmonary embolism.

Rivaroxaban is a reversible direct factor Xa inhibitor with peak plasma concentrations at 3 hours after ingestion. The half-life is 4–9 hours (up to 13 hours in patients age 65 years and older). Rivaroxaban is 60% renally cleared, but no dose reductions are required for mild renal insufficiency. Rivaroxaban is contraindicated with CrCl <30 mL/min. Rivaroxaban was investigated in 4 large phase III trials for prevention of VTE after total hip and knee arthroplasty (RECORD 1–4). All studies included older patients but excluded patients with renal and hepatic insufficiency (see Table 55–5). In all 4 trials, rivaroxaban 10 mg orally daily was superior to enoxaparin for a composite of total VTE and all-cause mortality. There were no significant differences in the rates of major bleeding or hepatic enzyme elevations between the 2 treatments. In the ROCKET study, approximately 14,000 patients with nonvalvular atrial fibrillation were randomly assigned to rivaroxaban 20 mg orally daily or dose-adjusted warfarin. Notably, patients with CrCl 30–50 mL/min were included in the trial with a dose reduction to 15 mg orally daily. Rivaroxaban was noninferior to warfarin for the prevention of stroke or systemic embolism (HR: 0.88; 95% CI: 0.74–1.03; P <0.001 for noninferiority). There was no significant between-group difference in the risk of major bleeding, although intracranial (0.5% vs. 0.7%, p = 0.02) and fatal bleeding (0.2% vs. 0.5%, p = 0.003) occurred less frequently in the rivaroxaban group. Rivaroxaban has been shown to be noninferior to warfarin for the treatment of acute VTE, but it has yet to be approved for this indication.

Although the characteristics of a more rapid onset of action, and a more predictable anticoagulant effect make these newer agents an attractive alternative to warfarin, caution is still needed when used in older patients. The convenience of no regular monitoring of coagulation also means there is no mechanism to objectively assess adherence to therapy. This may be more problematic for older patients in whom a fixed-dose regimen may not universally apply because of their variations in renal function and body weight, where the safety and efficacy of these agents is uncertain. Additionally, the lack of monitoring may potentially lead to missed opportunities for early detection of a complication because of lack of regular patient–provider interaction.

Bleeding is the primary complication of anticoagulation therapy. Older patients are particularly susceptible to bleeding complications on anticoagulation as a result of their inherent risk for falls; comorbidities such as renal failure, hepatic dysfunction, malnutrition, malignancy, amyloid angiopathy; concomitant use of antiplatelet agents; and noncompliance with drug regimens. Although reversal of older therapeutic agents such as UFH and warfarin is possible, many of the newer anticoagulants, including LMWHs, fondaparinux, parenteral direct thrombin inhibitors, and the novel oral anticoagulants, do not have a complete and specific antidote. Therefore the ideal method to manage bleeding with patients receiving these therapies is not known. Furthermore, accurate and widely available laboratory tests to measure anticoagulant activity may not be available for these newer agents. Although laboratory monitoring is not routinely required for patients on dabigatran or rivaroxaban, special clinical scenarios, such as clinically significant bleeding, may call for measurement of anticoagulant effect. At therapeutic doses, dabigatran prolongs the thrombin time/thrombin clotting time (TT/TCT) and activated partial thromboplastin time (aPTT), and has little effect on the prothrombin time (PT). Even though the TT/TCT and aPTT are the most effective and widely available coagulation assays to determine dabigatran activity, the therapeutic range of these tests is not well defined, and these tests are best used to determine the presence or absence of the drug. Rivaroxaban causes prolongation of both the PT and aPTT but exhibits greater sensitivity for PT. However, PT prolongation is not specific. Anti-factor Xa assays specific for rivaroxaban would be ideal for determining plasma rivaroxaban concentrations. Clinicians should not routinely use these laboratory tests to monitor and adjust dabigatran doses or assess the degree of bleeding risk for surgical procedures. However, in an emergency a normal TT/TCT and aPTT rules out the presence of significant amounts of dabigatran. Similarly, a normal PT and aPTT should indicate negligible amounts of rivaroxaban concentrations. Suggestions for the management of bleeding complications in patients on anticoagulation are described in Table 55–6.

Interruption of anticoagulation before interventions with high-risk of bleeding must be weighed carefully against thrombotic risk. Renal and hepatic impairment, which can prolong clearance of anticoagulants, as well as the long half-life of several of these drugs, need to be considered prior to discontinuation of the drug. Recommendations for interruption of the newer anticoagulants, including LMWHs, fondaparinux, parenteral direct thrombin inhibitors, and the new oral anticoagulants, are available. Restarting anticoagulation in a timely manner postoperatively depends on individualized assessment of risks of bleeding from the procedure and thrombosis for underlying hypercoagulable state. Postoperatively, it is critical to note that, unlike warfarin, the newer anticoagulants have more immediate onsets of action. Therefore, if these drugs are interrupted for surgery, they should not be reintroduced until hemostasis is assured.

Anticoagulants are among the most common drugs used to prevent and treat thrombotic and vascular disorders prevalent in the geriatric population. Special attention must be paid to the unique characteristics of older patients that may affect type, dose, monitoring, and management of bleeding of the anticoagulant chosen. Randomized controlled trials specifically addressing geriatric patients are needed to make evidence-based recommendations on the use of anticoagulants in this population.