By Karen Blum

Patient preferences, anticoagulant benefit versus risk and limited life expectancy should be considered when managing anticoagulation treatment at the end of a patient’s life.

“We have a lot of guidelines that cover when to start anticoagulants, but they’re conspicuously lacking for when and how to stop them,” said Anna Parks, MD, an assistant professor and the director of thrombosis at the University of Utah, in Salt Lake City. Decisions can be challenging because studies that target this patient population “are few and far between,” Dr. Parks said in a presentation at the 2024 American Society of Hematology Annual Meeting & Exposition, in San Diego.

“Patients’ priorities toward the end of life often diverge from what’s typically measured in studies, like thrombosis and bleeding,” she said. “There’s added difficulty in knowing how to match the care that’s available even after you successfully elicit priorities.”

There are three tenets to keep in mind when managing anticoagulants in the end-of-life setting, Dr. Parks said:

Recognize the prevalence of limited life expectancy in patients taking anticoagulants. While end of life is defined as life expectancy of one or fewer years, estimating prognosis can be difficult or imprecise, she said. Atrial fibrillation (AF) and venous thromboembolism (VTE) are diseases of aging, and chronologic age, functional status, comorbidities and polypharmacy are equally strong predictors of mortality (J Am Geriatr Soc 2019;67[4]:665-673). The ePrognosis website is a free online repository of evidence-based prognostic indexes that clinicians can use for a rough estimate of life expectancy (bit.ly/3DPyYIv).

image

A significant proportion of patients on anticoagulants—1 in 4 patients with AF and 2 in 3 with VTE—die within a year of diagnosis (Ann Intern Med 2022;175[8]:1065-1072), Dr. Parks said. Yet anticoagulants are continued for many patients up until death. For example, nearly one-third of patients with AF remained on anticoagulation during the last six months of life (JAMA Intern Med 2021;181[8]:1121-1123), and markers of more severe dementia and functional decline were associated with an increased risk for continuing treatment. In a study of patients with cancer-associated thrombosis, around half continued anticoagulation until death (BMJ Support Palliat Care 2022;12[e6]:e834-e837).

However, randomized controlled trials of anticoagulants often exclude patients with life expectancies of less than three to six months and those with poor functional status, said Dr. Parks, so their results—and guidelines that incorporate such results—may be less applicable to patients nearing the end of life (see sidebar).

Lessons From Anticoagulation Data

Although clinical trials on anticoagulants often exclude patients with limited life expectancy, some lessons can be gleaned from older populations, said Anna Parks, MD, an assistant professor and the director of thrombosis at the University of Utah, in Salt Lake City.

In atrial fibrillation, studies have shown the net clinical benefit of warfarin declines after 87 years of age, and the net clinical benefit of apixaban declines after age 92 (Circ Cardiovasc Qual Outcomes 2019;12[11]:e006212). “As you’re thinking about prescribing, stopping or continuing anticoagulants, you should consider the competing mortality risks,” she said.

For venous thromboembolism (VTE), data from the RIETE registry in Europe demonstrated that among patients aged 90 years and older, 1 in 5 died within 90 days of diagnosis. Among patients who presented with a pulmonary embolism (PE), fatal PE or VTE recurrence risk was 4% and overall bleeding rate was about 4.5 times VTE recurrence (J Am Geriatr Soc 2024;72[1]:113-125). However, this was reversed in patients who presented initially with deep vein thrombosis (DVT), for whom the risk of bleeding far exceeded VTE recurrence.

Hospice also presents a scenario from which to extrapolate, Dr. Parks said. One French study of hospice patients (most with cancer) found the incidence of bleeding to be 10%, with a case fatality rate of 20% (J Thromb Haemost 2017;15[3]:420-428). Risk factors for bleeding included anticoagulant use. Symptomatic DVT incidence was 0.5%.

“Although anticoagulation is started or continued in a substantial proportion of hospice patients and patients with limited life expectancy, there’s potentially uncertain benefit,” she said.

—K.B.

Incorporate the concepts of time to benefit and time to harm into anticoagulant decision-making. A long time to benefit (the time until a medication’s effect is evident) may exceed life expectancy for some patients, Dr. Parks said. If drugs have a short time to harm (the time required for a therapy to cause harm), patients with limited life expectancy may experience those harms and die before accruing the benefits.

“Bleeding risk from anticoagulants starts essentially as soon as you start the medication, and it’s highest with initiation,” she said. Whether to prescribe anticoagulation should consider a patient’s life expectancy and whether it exceeds expected time to benefit, versus risk factors for bleeding and time to bleeding risk.

Develop an individualized framework for starting, stopping or continuing anticoagulants based on patient priorities. As patients approach the end of life, outcomes measured in anticoagulant trials may become less relevant, Dr. Parks said. “There’s a ton of diversity in the type of care that patients want near the end of life that you don’t know about until you ask,” she said.

Patients may accept any treatment to prolong life, or trade-offs in quality of life for more time, while others prize quality of life above survival, she said. “The decision to start, continue or stop anticoagulants near the end of life needs to be personalized and reassessed.”

One way to improve decision-making is to use evidence-based communication tools such as a patient priorities care framework (JAMA Intern Med 2019;179[12]:1688-1697), in which clinicians help patients clarify their values and goals and state their care preferences. Next, consider treatment options with patients and family members within the context of patient priorities, incorporating functional status, bleeding and thrombosis risk, symptoms, and burdens. Then, align anticoagulation care accordingly, keeping in mind considerations such as preferred mode of administration. For example, because of nausea or pill burden, some patients may prefer parenteral agents.

A Pharmacist’s Take

There’s a lot that happens in end-of-life anticoagulation treatment conversations with patients, and Dr. Parks’ presentation “did a really nice job of summarizing and bringing to the forefront the most important parts,” commented Kelly Rudd, PharmD, BCPS, CACP, FCCP, a clinical associate professor of medicine at Oklahoma State University Center for Health Sciences, in Tulsa, in an interview.

Treatment is nuanced, she said, with some patients not wanting to become dependent on family, and altered drug pharmacokinetics as patients’ livers and kidneys start to shut down.

“Deciding when to continue or discontinue anticoagulants is a crucial team-based decision, with the patient at the center,” Dr. Rudd said. “Pharmacists play a key role in this process, and are uniquely positioned to integrate evidence-based literature, patient priorities and pharmacokinetic knowledge. This integration facilitates treatment conversations and guides anticoagulation stewardship near the end of life.”

The sources reported no relevant financial disclosures.