Atrial fib: apixaban outperforms rivaroxaban

 A recent large study of Medicare recipients with atrial fibrillation found that rivaroxaban was associated with more ischemic strokes as well as more hemorrhagic events than apixaban (see afib apix better than rivarox JAMA2021 in dropbox, or doi:10.1001/jama.2021.21222)

 

Details: 

-- retrospective cohort using the US Medicare database of those >65 years old, finding 581,451 patients with atrial fibrillation, newly prescribed rivaroxaban or apixaban, and followed for 4 years; study from 2013-2018

-- mean age 77, 50% female

-- CHA₂DS₂-VASc scores: mean 4.2 rivaroxaban, 4.4 apixaban

-- comorbidities: hypertension 90%, diabetes 35%, heart failure 30%, anemia 26%, COPD 22%, cancer 18%, GI or other bleeding 15%, CKD stage 3 in 16%

-- meds: ACE/ARB 62%, PPI 31%, loop diuretic 28%, oral steroids 25%, diltiazem or verapamil 21%, NSAIDs 17% (though they may have also been OTC), platelet inhibitors 16%

-- those on rivaroxaban had a more favorable prevalence of many risk factors for study outcomes versus apixaban (history of stroke or intracranial bleed 8.5% versus 9.8%; history of myocardial infarction 5.6% versus 6.8%; heart failure 27.9% versus 31.2%; acute kidney failure 7.6% versus 10.5%; at least stage 3 chronic kidney disease 13.8% versus 17.9%; history of bleeding 14.4% versus 15.6%; anemia 24.6% versus 27.1%; unintentional falls 9.5% versus 11.5%, and hospitalization in the 30 days before anticoagulation initiation 26.7% versus 29.0%). [ie. apixaban patients were pretty globally sicker and seemed to have higher risk of bleeding]

-- 227,572 patients were on rivaroxaban and 353,879 on apixaban

    -- 23% of each group were on low-dose meds (2.5 mg twice a day versus 5 mg twice a day for apixaban, 15 mg versus 20 mg once a day for rivaroxaban)

-- the study excluded patients who had potentially reversible causes of atrial fibrillation (e.g. thyrotoxicosis, alcohol), had mitral valve stenosis, or severe chronic kidney disease

 

--primary outcome was major ischemic or hemorrhagic events: ischemic stroke, systemic embolism, hemorrhagic stroke, other intracranial bleeding, and fatal extracranial bleeding (death within 30 days of bleeding onset)

-- secondary outcome: non-fatal extracranial bleeding and total mortality (which included fatal ischemic or hemorrhagic events leading to death within 30 days of event onset and other deaths during follow-up)

    -- in order to equalize these groups, they controlled for 208 covariates that could potentially be associated with both outcomes and for anticoagulant choice, an array of demographics, cardiovascular conditions (including claims-based components of the accepted CHA₂DS₂-VASc scores), and risk factors for bleeding. The mathematically adjusted rates of these outcomes are below.

 

-- A total of 474,605 person-years of follow-up with the median of 175 days

-- patients on rivaroxaban were more likely to discontinue the study drug or switch to a different oral anticoagulant than patients with apixaban: 33.4% versus 30.4% discontinued the meds and 7.0% versus 4.5% switched to a different anticoagulant

 

Results: 

-- major ischemic events (total events 6009, 14.6/1000 person-years) or hemorrhagic events (total 3807, 8.0/1000 person-years)

    -- rivaroxaban: 2838 patients, 14.8/1000 person-years, with adjusted rate of 16.1/1000 person-years

    -- apixaban: 4108 patients, 14.5/1000 person-years, with adjusted rate of 13.4/1000 person-years

        -- adjusted rate difference: 2.7 (1.9-3.5), 18% more with rivaroxaban, HR 1.18 (1.12-1.24)

 

-- adjusted rates for primary outcomes, comparing rivaroxaban to apixaban:

    -- ischemic event: 8.6 versus 7.6/1000 person-years, rate difference 1.1, HR 1.12 (1.04-1.20)

    -- ischemic stroke: 8.3 versus 7.2/1000 person-years, rate difference 1.1, HR 1.12 (1.05-1.21)

    -- systemic embolism: 0.4 versus 0.3/1000 person-years, rate difference 0, nonsignificant

    -- hemorrhagic event: 7.5 versus 5.9/1000 person-years, rate difference 1.6, HR 1.26 (1.16-1.36)

    -- other intracranial hemorrhage: 3.5 versus 3.2/1000 person-years, rate difference 0.3, HR 1.09 (0.98-1.22)

    -- fatal extracranial bleeding: 1.4 versus 1.0/1000 person-years, rate different 0.4, HR 1.41 (1.18-1.70)

 

-- adjusted rates for secondary outcomes, comparing rivaroxaban to apixaban:

    -- nonfatal extracranial bleeding: 39.7 versus 18.5/1000 person-years, rate difference 21.1, HR 2.07 (1.99-2.15)

    -- gastrointestinal bleed: 35.2 versus 16.3/1000 person-years, rate difference 19.0, HR 2.09 (2.01-2.18)

    -- total mortality: 44.2 versus 41.0/1000 person-years, rate difference 3.1, HR 1.06 (1.02-1.09)

    -- fatal ischemic or hemorrhagic event: 4.5 versus 3.3/1000 person-years, rate difference 1.2, HR 1.34 (1.21-1.48)

 

-- dose of medications given: 23% in each group were on reduced doses, especially if they were older (83 versus 75 yo), were more likely to be women (63% versus 46%) and had greater prevalence of other risk factors for stroke (CHA₂DS₂-VASc scores 5.0 versus 4.1) and bleeding

    -- even at reduced doses, rivaroxaban had increased risk for primary outcome (adjusted rate 27.4 versus 21.0 per 1000 person-years, rate difference 6.4, HR 1.28 (1.16-1.40)

    -- patients receiving reduced dose rivaroxaban had greater risk for nonfatal extracranial bleeding, whereas total mortality did not differ from those on apixaban

    -- those on standard dose meds still had increased risk for primary outcome with rivaroxaban, adjusted rate 13.2 versus 11.4/1000 person-years, rate difference 1.8, HR 1.13 (1.06-1.21)

 

-- No significant difference in outcomes by excluding those with cardiac valve replacements, stage 3 CKD, or extreme propensity scores (meaning much more mathematical correction)

    -- and, given that medication nonadherence increases over time, restriction of the follow-up to one year after initiation of the med and reduction of the allowable gap in drug supplied to the patients to 7 days instead of 30 did not materially affect the results

 note that these curves began to separate after only 6 months

Commentary: 

-- as we all know, atrial fibrillation is remarkably common in the US. Estimates are that there are 3 to 6 million people currently and likely to reach 6 to 16 million by 2050

-- stroke can certainly be devastating, and the risk increases roughly 5-fold with atrial fibrillation; and atrial fibrillation explains about 15% of all strokes

-- Direct-acting oral anticoagulants (DOACs) have largely replaced vitamin K antagonists (VKAs) as the treatment of choice for atrial fibrillation, given their presumably more predictable pharmacokinetics, ease-of-use, and fewer drug interactions than VKAs

-- there are 4 licensed DOACs in the US, with rivaroxaban and apixaban being the leading 2 by far, and they are prescribed more frequently than VKAs such as warfarin

 

-- This very large observational study found that despite rivaroxaban being given to healthier patients at lower risk of ischemic events and bleeding, the adjusted differences for the two medications very strongly favored apixaban for the primary and secondary outcomes overall, as well as essentially every individual component of these different outcomes, including total mortality and fatal ischemic or hemorrhagic events

-- additional findings in the study were that even those on reduced doses of medications (23% of the patients) because of anticipated greater susceptibility to both anticoagulant efficacy and safety, there still was an increased risk for those on rivaroxaban even at this lower dose

 

-- Because rivaroxaban and apixaban actually have very similar half-lives (though the former is only once a day and the latter twice a day…), rivaroxaban does have much greater peak-to-trough blood concentration differences. And one argument to explain both the increased clotting events as well as the increased bleeding events with rivaroxaban is the increased peak concentrations at times (leading to bleeding) as well as both a wear-off effect leading to more clotting and the more exaggerated potential effect by missing a dose of rivaroxaban (other studies have found pretty clearly that adherence to these medications is suboptimal, and missing one dose of rivaroxaban maybe expose people to much higher risk of clots versus missing a dose or two of apixaban)

 

-- another recent article, also analyzing retrospective cohort studies in a commercial health care insurance database, used propensity-matched cohorts of 36,236  patients and compared apixaban and rivaroxaban for venous thromboembolic disease (VTE) treatment, finding also that apixaban did better: significantly lower rates of recurrent VTE (23% decrease, risk difference at 2 months was 0.6% and at 6 months at 1.1%) and lower risk of GI or intracranial bleeding requiring hospitalization (risk different at 2 months 1.1% and at 6 months 1.5% ), see https://www.acpjournals.org/doi/10.7326/M21-0717 )

 

-- As a background, there were several drug company shenanigans involved with DOAC development:

    -- the use of defective point-of-care INR machines in the rivaroxaban trial, which altered the risk-benefit analysis in its comparison with warfarin (see http://gmodestmedblogs.blogspot.com/2016/02/rivaroxaban-is-it-really-better-than.html for a review of the BMJ article that exposed this)

    -- another series of articles in the BMJ that exposed pretty blatant drug company malfeasance with regard to dabigatran, the first DOAC: see http://gmodestmedblogs.blogspot.com/2014/07/big-pharma-yet-againagaindabigatran.html and http://gmodestmedblogs.blogspot.com/2014/11/dabagratan-again.html

    -- the bottom line: the drug companies had a vested interest in presenting an alternative to warfarin, a drug that did not need monitoring (though some of the above blogs suggest there was impressive data that they should require drug levels, and considered promoting the lack of need was in fact dangerous….) and that these meds were easy to take (I suspect that the reason that rivaroxaban is a once a day drug, though it has the same half-life as apixaban, is that that would be a selling point. Why it’s almost like the drug companies are more interested in $$$$ (with lots of accompanying zeros) than in providing optimal products to help people????

 

-- one persistent concern I have with DOACs is the reversal agents, in case someone is in an accident, or they just get significant internal bleeding: both are coagulation factor Xa inhibitors, and they need the same reversal agent andexanet alfa. However, it is unclear how available that is globally, especially in resource poor countries or even in rural areas in the US. This is a major concern for me given that the vast majority of my patients are from such countries and travel back and forth from them. And vitamin K is available....

 

Limitations:

-- the rivaroxaban group was clearly at lower risk of ischemic and hemorrhagic complications at the outset; mathematical modeling did attempt to equalize the groups, but there may well still be unsuspected residual confounding that was not taken into account

    -- of note, the overall differences in patient population between the two medications was typically quite small, suggesting that there was not much channeling of patients to one medication over the other
-- as with all observational studies there could be unmeasured factors that might have influenced the outcome, including differences in patient or clinician preferences. Perhaps clinicians preferentially chose rivaroxaban because they were concerned about patient lack of adherence to a twice a day medication as with apixaban, and this may reflect significant other differences between the populations on these different medications, limiting the validity of the findings

-- the data is from large databases and relies on computer-based coding of information. this is always fraught by potential inconsistencies in clinician data input

-- the study only involved people over 65 years old who are on Medicare parts A and B as well as part D. Medicare advantage programs (part C) enrollees were not included, and those who had not signed up for all of these components of Medicare. and no younger patients. These would limit generalizability to the broader population

 

So, pretty impressive study overall suggesting that in cases where a DOAC is chosen, apixaban has really impressive advantages over rivaroxaban, both for preventing ischemic events and minimizing the potential bleeding risk. 

geoff

 

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@uphams.org

 

to get access to all of the blogs (2 options):

1. go to http://gmodestmedblogs.blogspot.com/ to see them in reverse chronological order

2. click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

3. or you can just click on the magnifying glass on top right, then  type in a name in the search box and get all the blogs with that name in them

 

or: go to https://www.bucommunitymedicine.org/ , a website from the Community Medicine section at Boston Medical Center.  This site does have a very searchable and accessible list of my blogs (though there have been a few that did not upload over the last year or two). but overall it is much easier to view blogs and displays more at a time.

 

 

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list