News · Heart & Metabolic
After a heart valve replacement, a blood thinner beat aspirin by nearly half
Aspirin is the default after a valve is fitted through a catheter, on the reasoning that anything stronger bleeds. Randomizing 360 patients found the stronger drug both worked better and did not bleed more.
- Patients on the anticoagulant had 16% of events against 29% on aspirin.
- Bleeding, the reason aspirin is preferred, was not higher on the stronger drug.
- The valve is fitted through a catheter rather than by open surgery.
- 360 patients, average age 74, randomized.
- A single trial of this size settles nothing on its own.
There is a particular kind of medical decision that gets made once, on reasonable grounds, and then stops being examined.
After a replacement aortic valve is threaded into place through an artery, the patient needs something to stop clots forming on the new hardware. The choice has been aspirin, and the reasoning is straightforward: these patients are in their seventies and eighties, stronger blood thinners cause bleeding, and bleeding in the elderly is dangerous.
Writing in JAMA, researchers tested whether the cautious choice was the right one. Of the 360 participants randomized, half received a direct oral anticoagulant instead.
The efficacy result
Events occurred in 27 patients on the anticoagulant against 48 on aspirin. In percentages, 16.2% against 28.6%.
That is a reduction of roughly 45%, with a range that stays comfortably clear of no effect.
For a group where the alternative was aspirin chosen largely by convention, a gap that size is not a refinement. It suggests the convention was costing patients events.
The result that removes the objection
Efficacy alone would not settle anything, because the entire case for aspirin is about safety rather than effectiveness. Everybody expected the stronger drug to work better; the question was what it cost.
The primary safety end point occurred in 13 patients on the anticoagulant and 19 on aspirin, 7.5% against 10.6%.
Fewer bleeds on the drug that was supposed to bleed more. The range around that difference is wide, running from a meaningful advantage to a modest disadvantage, so the honest statement is that the trial found no bleeding penalty rather than that it found a bleeding benefit.
Either way, the premise the standard practice rests on did not hold up here.
Why the procedure matters to the question
Aortic stenosis, where the valve stiffens and narrows, is the commonest valve problem of old age. Heart valve diseases once meant open-heart surgery, which many older patients could not survive.
The catheter approach changed that, and it changed who gets treated. The population now receiving new valves is older and frailer than the surgical population ever was, which is exactly the population in whom bleeding fears are most acute and in whom under-treatment is easiest to justify.
That makes the default worth interrogating, and it is why a trial in 360 people at an average age of 74 is worth more than its size suggests.
What 360 patients cannot do
The efficacy difference is large enough to be visible at this scale. The safety comparison is not.
Serious bleeds are uncommon events, and a trial with 32 of them across both arms cannot exclude a real excess of the size that would change practice. The confidence range spans from 9.5% fewer to 2.8% more, which is another way of saying the trial is underpowered for the question that matters most.
There is also the ordinary caution about a single trial. Practice changes on replication, and one result of this kind is a reason to run the larger version rather than to rewrite the protocol.
Who this does not apply to
Patients who already take an anticoagulant for atrial fibrillation are not in this comparison; they were always going to be anticoagulated. The relevant group is the one who would otherwise have been sent home on aspirin alone, which is most of them.
The wider point
Defaults in medicine accumulate. Somebody makes a sensible call in the absence of evidence, it becomes the protocol, the protocol becomes the answer to a question nobody is asking any more, and the trial that would test it never gets funded because the practice is already established.
This is what it looks like when someone runs it anyway. The cautious option turned out to be the one with worse outcomes and no safety advantage, in the group it was designed to protect.
People also ask
What did the trial find?
The primary efficacy end point occurred in 27 participants (16.2%) on the anticoagulant and 48 (28.6%) on aspirin (risk ratio 0.55; 95% CI 0.37-0.82; P = .004). The primary safety end point occurred in 13 (7.5%) versus 19 (10.6%), a risk difference of -3.3% (95% CI -9.5% to 2.8%).
What is this procedure?
Transcatheter aortic valve implantation, or TAVI: a replacement aortic valve threaded up through an artery rather than fitted during open-heart surgery. It has largely replaced surgery for older patients.
Why has aspirin been the default?
Because the patients are old, and stronger blood thinners cause more bleeding in old people. Aspirin was the cautious choice, and cautious choices become standard without always being tested.
What is the significance of the bleeding result?
It removes the reason for the caution. The whole argument for aspirin rests on the stronger drug bleeding more; here the bleeding numbers favored the anticoagulant, though the range around that difference is wide enough to include either.
Is 360 patients enough?
For the efficacy result, the difference was large enough to be clear. For the safety result it is not: bleeding events are uncommon, and a trial this size cannot rule out a meaningful excess.
Does this apply to everyone after the procedure?
No. Patients with atrial fibrillation already receive an anticoagulant for that reason. This addresses the group who would otherwise be on aspirin alone.
Should anyone change their medication?
No. What follows a valve procedure is decided by the team that performed it, weighing bleeding risk that this trial could only partly measure. This is general information rather than medical advice.