News · Brain & Mental Health
A blood thinner outperformed a standard migraine drug in people with a hole in the heart
Roughly one adult in four has a small flap left open between the heart's upper chambers, and it is more common in people with migraine. Randomizing 984 of them put an anticoagulant well ahead of the usual preventive.
- Rivaroxaban halved migraine days for 78% of patients, against 62% on the standard drug.
- Aspirin and clopidogrel matched the standard drug rather than beating it.
- Everyone in the trial had a small opening between the heart's upper chambers.
- Twelve weeks is short for a condition managed over years.
- 984 adults across 39 hospitals in China, with treatment assigned at random.
About a quarter of adults are walking around with a small flap between the upper chambers of the heart that never sealed after birth. It causes no symptoms in most of them and is usually found by accident.
It is also noticeably more common in people who get migraine with aura, which has produced two decades of argument about whether closing it helps. Trials of surgical closure have been mostly disappointing, and the question has stayed open.
Researchers writing in The BMJ approached it from a different direction. Rather than closing the hole, they asked whether treating the blood passing through it changes anything, enrolling 1000 adults aged 18-64 with confirmed openings and frequent migraine across 39 hospitals, of whom 984 reached the analysis.
Four drugs, one comparison
Participants were assigned to one of four treatments for twelve weeks: aspirin, clopidogrel, rivaroxaban, or metoprolol.
The first three thin the blood or stop platelets clumping, by increasingly powerful mechanisms. The fourth, metoprolol, is a beta blocker with no effect on clotting at all, and it is one of the standard drugs used to prevent migraine.
Using metoprolol as the comparison rather than a placebo is a deliberate choice. It answers the question a clinician actually faces, which is what to prescribe, rather than the question of whether anything works better than nothing.
What separated the arms
All three antithrombotic drugs matched metoprolol. Aspirin, clopidogrel, and rivaroxaban were all non-inferior on the primary endpoint, which was the share of patients whose monthly migraine days or attacks fell by at least half.
Then one pulled clear. Responder rates were 61.7% with aspirin, 66.8% with clopidogrel, 78.4% with rivaroxaban, and 61.8% with metoprolol.
Roughly four patients in five improving by half or more, against three in five on the standard drug. That is a wide margin for a migraine trial, where preventive treatments typically deliver modest gains.
The gradient that supports the theory
The three blood-affecting drugs did not perform identically, and the order they fell in is the interesting detail.
Aspirin, the weakest antiplatelet, did no better than the beta blocker. Clopidogrel, a stronger antiplatelet, sat slightly higher. Rivaroxaban, which blocks the clotting cascade itself rather than platelet stickiness, was well ahead.
If the opening lets small clots or clot-associated chemicals bypass the lungs and reach the brain, then a drug that acts further upstream should do more. The results line up with that prediction, which is not proof of the mechanism but is the pattern the mechanism predicts.
Why the design needs stating plainly
This was an open-label trial: everyone knew which drug they were taking. In a condition measured by patient-reported headache diaries, that matters, because expectation shapes reporting.
The outcome assessment was blinded, which limits the damage without removing it. A patient who knows they are on the novel anticoagulant rather than the familiar beta blocker may record their headaches differently.
Twelve weeks is also short. Migraine is managed over years, migraine frequency fluctuates on its own, and a three-month window catches some of that natural variation.
The risk nobody should skip past
Rivaroxaban is an anticoagulant. Migraines are severe headaches that can be disabling, and they are not usually life-threatening; serious bleeding is.
Long-term anticoagulation is accepted where the alternative is a stroke. Accepting it to reduce headache days is a different calculation, made by a different kind of patient, and it needs bleeding data over years rather than over a quarter.
The trial recorded bleeding among its safety outcomes. Twelve weeks in a thousand people is not enough to characterize uncommon serious bleeds.
What it actually establishes
That in people with migraine and a patent foramen ovale, a drug acting on the clotting system outperformed a standard preventive over three months, in a single-country open-label trial.
That is a genuine result and a narrow one. It applies to a specific group who have had an echocardiogram showing a specific finding, and it needs a blinded trial running long enough to weigh headaches against bleeds before it could sensibly change anyone’s prescription.
What it does do is move a twenty-year argument forward. The question was whether the hole matters for migraine. This is the strongest evidence yet that something passing through it does.
People also ask
What did the trial find?
For the primary endpoint, aspirin, clopidogrel and rivaroxaban were all non-inferior to metoprolol. Responder rates for a 50% or greater reduction in monthly migraine days or attacks at weeks 9-12 were 61.7% with aspirin, 66.8% with clopidogrel, 78.4% with rivaroxaban and 61.8% with metoprolol, with rivaroxaban showing a statistically higher rate.
What is a patent foramen ovale?
A small flap between the upper chambers of the heart that everyone has before birth and that usually seals shortly after. In roughly a quarter of adults it stays open. Most people never know, and it is more common among people with migraine with aura.
Why would a blood thinner help migraine?
The leading idea is that small clots or chemical triggers normally filtered by the lungs pass through the opening into the arterial circulation and reach the brain. Reducing clot formation would reduce that traffic. The trial tested the effect, not the explanation.
How were the drugs compared?
Against metoprolol, a beta blocker widely used to prevent migraine, rather than against placebo. That makes the comparison clinically useful and means none of the arms establishes an effect over no treatment at all.
Does everyone with migraine have this?
No. The trial enrolled only people with a confirmed opening and at least four migraine days a month, so the result applies to that group and not to migraine generally.
What are the risks?
Rivaroxaban is an anticoagulant and raises bleeding risk, which is the trade-off in any long-term use. Bleeding was among the trial's safety outcomes, and twelve weeks is short for detecting uncommon serious bleeds.
Should anyone ask for this?
No. This is one open-label trial in one country and it does not change prescribing. Migraine prevention should be discussed with a clinician who knows the individual history. This is general information rather than medical advice.