It is well established that older individuals with atrial fibrillation (AF) are less
likely to receive oral anticoagulant (OAC) therapy compared with their younger counterparts,1,
2 and when treated with vitamin K antagonists (VKAs), there is a relatively high rate
of discontinuation resulting in a high rate of stroke or death.3, 4 This undertreatment
of the very elderly represents a paradox because older patients are at higher risk
of stroke and are more likely to need anticoagulant therapy compared with younger
patients.5, 6 Why are physicians reluctant to prescribe therapy? There are many reasons
for the undertreatment of AF in the elderly including physician‐related factors, patient‐related
factors, and the practical aspects of therapy.7 An overriding concern, however, is
the fear of putting the patient at risk for major bleeding as a result of anticoagulant
therapy, while the fear of leaving the patient open to stroke is of lesser concern.4,
7, 8, 9 This is sometimes expressed as the fear of creating a sin of commission versus
a sin of omission (by doing something we should not do versus not doing something
we should do).10 The consequence of the latter (ie, stroke) is chalked up to the natural
course of the disease. Until recently, studies to address the net benefit of anticoagulant
therapy for stroke prevention in AF in the very elderly were lacking. Now we have
substantial evidence in progressively older cohorts to put this issue to rest.
In elderly patients with AF, numerous trials have documented the increased risk of
stroke while also showing an increased risk of bleeding with OAC therapy.5, 6 Until
recently, only therapy with the VKAs was available where many factors come into play
in how patients fare, including the all‐important system of dose management of this
complex drug to keep the patient in therapeutic range.11 But even in the best of settings,
such as randomized clinical trials or anticoagulation clinics, major bleeding is increased
in the elderly when taking anticoagulants. Consequently, physicians may withhold therapy
or prescribe aspirin, a clearly less effective antithrombotic, but one that physicians
feel is less likely to cause major bleeding. Studies now show this to be a fallacy
when applied to the elderly. Until recently, there was a dearth of studies that included
“very old” patients, generally ≥80 or 85 years of age, but this void is quickly being
filled.
In this issue of JAHA, Patti et al12 provide evidence favoring treatment with anticoagulant
therapy to prevent stroke or systemic embolism in the very elderly that outweighs
the risk of major bleeding. They report on a subanalysis of the PREFER in AF Registry
(PREvention oF thromboembolic events‐European Registry in Atrial Fibrillation), a
registry of over 7000 consecutive patients with AF from 461 centers in 7 European
countries conducted between 2012 and 2014.13 Although the report has limitations as
a prospective registry, it has strength reflecting real‐world antithrombotic therapy
and it reports on a sizable number of very old patients ≥85 years of age (505) and
compares them with 5907 patients <85 years. As expected, the older cohort was at greater
risk for stroke and bleeding with more comorbidities and higher CHA2DS2VASc score
than the younger cohort.
In all patients, those treated and not treated, the occurrence of stroke/transient
ischemic attack/systemic embolism was substantially higher in the older than in the
younger cohort (4.8% per year versus 2.3% per year, respectively; P=0.0006). Similarly,
the older cohort had a higher rate of major bleeding than their younger counterparts,
although this did not reach significance (4% per year versus 2.7% per year, respectively;
P=0.11), unless those ≥85 years were compared with those <75 years where the rate
was 1.9% per year; P=0.001. Fifty‐one percent of all major bleeds were gastrointestinal,
9% intracerebral, and 43% other sites of bleeding. When comparing those treated with
anticoagulants versus not treated or treated only with antiplatelet agents, there
was a favorable odds ratio for reduced stroke/transient ischemic attack/systemic embolism
in both the elderly cohort (0.64; P=0.37) and in those <85 years (0.74; P=0.26). This
represented a greater absolute reduction of 2% in the elderly compared with 0.5% in
those <85 years. In those ≥90 years of age, the absolute risk reduction was even greater
at 4.6%. What is most important is the finding that major bleeding was not significantly
different in either cohort between those treated compared with those not treated with
anticoagulants; in the elderly, 4.2% versus 4.0%, respectively; P=0.77; in those <85 years,
3.4% versus 2.9%, respectively; P=0.74. Similar findings were seen in an exploratory
analysis of those ≥90 years. When comparing those on anticoagulant therapy versus
those only on antiplatelet therapy in the very elderly, the incidence of major bleeding
was similar (4.1% versus 3.9%, respectively), but higher than in those without any
antithrombotic therapy (4.1% versus 2.8%). Finally, net clinical benefit (thrombosis+bleeding+myocardial
infarction) significantly favored treatment in the very elderly (P=0.036). Although
not the focus of this report, an incidental finding is that 78% of the very elderly
were receiving anticoagulants and over 80% of the entire cohort with a CHA2DS2VASc
score ≥2 received anticoagulants, thus documenting the changing nature of contemporary
physician prescribing behavior. This suggests a greater willingness to treat patients
with AF even in the very old compared with studies from the 1990s and early 2000s.
This trend has been seen in other recent registries as well14 and indicates an overall
improvement in complying with professional guidelines. The older cohort, however,
was also treated more frequently with antiplatelet drugs compared with younger patients
(15% versus 11%, respectively), a therapeutic intervention that is both less effective
and not any safer.
What are the take‐home messages from this prospective observational registry? First,
it focuses on the very elderly and confirms the higher risk of stroke and major bleeding
compared with a younger cohort. Second, it shows that the benefit of treating such
patients with anticoagulants outweighs the risk of major bleeding. Third, it supports
evidence that antiplatelet therapy with aspirin is associated with a similar risk
of major bleeding as with OAC. Fourth, and as an incidental finding, it suggests that
contemporary physician prescribing patterns for stroke prevention in AF in the very
elderly are changing, with a greater willingness to anticoagulate such patients.
These results support other recent studies of anticoagulant therapy in the very elderly.15,
16, 17, 18 The BAFTA trial (Birmingham Atrial Fibrillation Treatment of the Aged Study)
compared VKA therapy with antiplatelet therapy as stroke prevention in 973 patients
75 years or older.15 Investigators showed a significant reduction in stroke, systemic
embolism, or intracranial hemorrhage with VKA therapy (1.8% versus 3.8%; relative
risk 0.48, 95% CI 0.28–0.80; P=0.003). At the same time, extracranial hemorrhage was
no different between groups (VKA 1.4% versus aspirin 1.6%). In a very large cohort
of 4093 patients ≥80 years of age treated with OAC for a variety of indications, Poli
et al18 found a major bleeding rate of only 1.87%, which compares very well with the
bleeding rate in the BAFTA trial. Ogilvie et al,19 in a meta‐analysis of outcomes
of OAC therapy compared with aspirin therapy in real‐world patients outside of clinical
trials, further confirmed the benefit of OAC and the lack of improved safety with
antiplatelet therapy.
What do these findings mean in the age of the new direct oral anticoagulants (DOACs),
which are rapidly gaining market share for stroke prevention in AF? In the Patti et al
trial,12 DOACs had only a small penetration (approximately 6% in each group), limiting
an analysis of their performance in the elderly. But in subanalyses of elderly cohorts
in the phase 3 AF trials of the DOACs compared with warfarin, the effectiveness and
safety of DOACs compared with warfarin was maintained20, 21, 22 and potentially even
greater with similar safety when compared with aspirin.23 To be sure, the use of DOACs
in the very elderly requires special attention to a number of clinical issues such
as renal function, drug–drug interactions, tolerability, and others,24 but most of
these are easily manageable. Accordingly, the DOACs, with their improved safety performance,
have the potential to change the equation even further, providing a greater net clinical
benefit in the very elderly compared with the VKAs.
As other editorials on this topic have proclaimed,25, 26 it is time for all physicians
to recognize that although the very elderly with AF have a higher risk of stroke and
major bleeding than those who are younger, the increase in the risk of stroke is greater
than the increase in the risk of major bleeding, thus presenting the opportunity for
an even greater potential for stroke reduction versus a risk of major bleeding. Secondly,
antiplatelet therapy with aspirin is not only less effective, it is also no safer
than OAC therapy in the very elderly. It is gratifying to see from this registry that
physicians are more willing to employ anticoagulant therapy in these high‐risk patient
populations, suggesting that they are beginning to understand that acts of omission
are far more serious than acts of commission and that it is no longer a sin to treat
the very elderly with anticoagulant therapy.
Disclosures
Dr Ansell reports consultant activities and honoraria from Bristol Myers Squibb, Pfizer,
Daiichi Sankyo, Janssen, and Boehringer Ingelheim.