A dural arteriovenous fistula is an abnormal artery-to-vein connection in the lining of the skull that can cause pulsatile tinnitus; the type draining into veins on the brain's surface can bleed, and embolisation often cures the sound. In one series, the tinnitus disappeared completely in 21 of 26 people treated.
Of all the causes on the pulsatile tinnitus list, this is the one where the distance between the best and worst outcomes is widest. It is also the cause most likely to be missed, for a reason worth understanding.
What it is
A 2022 review in Stroke: Vascular and Interventional Neurology describes dural arteriovenous fistulas as abnormal arterial-to-venous connections contained within the dural leaflets — the tough membrane lining the inside of the skull. That containment is what makes them a distinct category from the arteriovenous malformations found inside brain tissue.
The functional problem is simple to state. Blood at arterial pressure is being delivered directly into a vein, skipping the capillary bed that normally steps the pressure down. The result is high-velocity turbulence a few centimetres from the inner ear, in bone that conducts it well.
That is why a dAVF produces a whooshing or rushing synchronised to the pulse rather than the steady tone of most tinnitus, and why it belongs to the same physical category as sigmoid sinus wall abnormalities and glomus tumours: a real sound, made near the ear, not a signal generated by the auditory system.
Why pulsatile tinnitus alone is enough to matter
A 2016 study in Scientific Reports reviewed 220 patients diagnosed with dAVF at a Korean centre between 2003 and 2014. Of those, 30 — 13.6% — presented with pulsatile tinnitus as their only initial symptom.
Not tinnitus plus headaches. Not tinnitus plus visual disturbance. Tinnitus, alone, as the entire presentation.
Where the fistulas sat, in that series:
- Transverse–sigmoid sinus — 70.0%, by a wide margin the commonest location.
- Hypoglossal canal — 10.0%.
- Middle cranial fossa — 6.7%.
The authors' conclusion is a direct instruction to clinicians, and it is the reason this article exists: because pulsatile tinnitus may be the only initial symptom in more than 10% of dAVF cases, not only otolaryngologists but also neurologists and neurosurgeons should evaluate patients with pulsatile tinnitus meticulously.
That sentence is worth carrying to an appointment, because the failure mode it describes is not a rare one. Pulsatile tinnitus arriving without headaches, without visual symptoms and without neurological signs looks reassuring. In this condition it is not evidence of anything.
The feature that separates a nuisance from a danger
Not every dAVF is a threat, and the thing that decides it is not the volume of the sound.
The 2022 review draws the line at cortical venous drainage — whether the fistula pushes blood backwards into the veins on the surface of the brain:
- Without cortical venous drainage, a dAVF may be found incidentally, or during a workup for pulsatile tinnitus or eye symptoms.
- With it, the review states that seizures, neurological deficits or even frank hemorrhage may occur, because normal venous outflow from brain tissue is disrupted and venous pressure rises behind it.
In the Korean series, drainage patterns split 73.3% sinus or meningeal against 26.7% sinus drainage with cortical venous reflux. Roughly a quarter of the fistulas found in patients whose only complaint was a noise in one ear were of the type carrying the higher-risk drainage pattern.
You cannot tell which you have from the sound. It is determined on imaging, and that is the entire argument for imaging.
The trap: it can look like the benign diagnosis
The article here on sigmoid sinus wall abnormalities argues that a finding on a scan is not automatically the cause of the sound, on the strength of a case-control study in which the abnormality showed up in 9% of people with no tinnitus at all. There is a sharper version of that caution, and it comes from a 2025 study in the Ear, Nose & Throat Journal.
The researchers looked at 80 people whose main symptom was pulsatile tinnitus and who had a dAVF confirmed on MR angiography — so, in every case, the answer was known. Then they asked what those patients' ordinary workups had shown.
Sigmoid sinus wall anomalies and jugular bulb anomalies were present in 40.8% of them.
Four in ten people with a fistula also had exactly the finding that gets read as a benign venous cause. Their conclusion, stated plainly: the presence of a sigmoid sinus wall anomaly on CT together with a positive jugular vein compression test should not be considered conclusive for diagnosing venous pulsatile tinnitus.
The same study offers two things that point the other way. A "moth-eaten sigmoid plate" appearance on non-contrast CT was found in 29 of 37 patients with a transverse–sigmoid fistula — a sensitivity of 78.4% (95% CI 0.63–0.89). And a positive retroauricular compression test, as opposed to a jugular one, was described as a strong indicator of dAVF.
Two notes on that, because both matter. A sensitivity of 78.4% means roughly one in five fistulas at that site did not show the sign, so its absence proves nothing. And the compression tests are examination manoeuvres performed by a clinician. Pressing on your own neck is not a home diagnostic, and compressing a jugular vein is not something to experiment with.
What treatment achieves
This is where the article turns, because the numbers are unusually good.
Of the 30 patients in the Korean series, 26 underwent transarterial embolisation — a catheter threaded up to the fistula and the connection blocked off from the arterial side.
- Tinnitus disappeared completely in 21 — 80.8%.
- Markedly improved in 3 — 11.5%.
- Unchanged in 2 — 7.7%.
The authors' conclusion: in most cases, pulsatile tinnitus originating from a dAVF can be cured with transarterial embolisation, regardless of location and venous drainage pattern.
Cured is a word that rarely applies to tinnitus. Can tinnitus be cured exists precisely because the honest answer for the ordinary kind is no. This is one of the narrow exceptions, and it exists because the sound has a physical generator that can be switched off — not because anything was done to the auditory system.
The 2022 review adds that for debilitating symptoms or for fistulas with cortical venous drainage, endovascular treatment via arterial or venous access is the usual approach, with surgical disconnection and radiosurgical obliteration as alternatives for appropriately selected lesions.
How it gets found
The 2024 review in Current Pain and Headache Reports sets out the pathway, and it is useful for knowing what to expect rather than for self-navigating.
CT and MRI have comparable diagnostic yield overall, though each is more sensitive to particular causes. If initial vascular imaging is negative and a vascular cause is still strongly suspected, digital subtraction angiography may be what settles the question — a catheter study, more invasive than a scan, and the reference standard for this specific problem.
The same review lists what else is on the differential: tumours, arteriovenous malformation or fistula, idiopathic intracranial hypertension, venous sinus stenosis, otosclerosis, patulous eustachian tube, and bony defects including superior semicircular canal dehiscence. Most of those have their own article, linked above, which is the shape of a proper differential rather than a list of scary words.
What to do with this
- If you have pulsatile tinnitus and have not had vascular imaging, that is the gap, and this article is the reason it is worth closing rather than deferring.
- If imaging found a sigmoid sinus abnormality, it may well be the answer — and the 2025 study is a fair thing to raise, because 40.8% of confirmed fistula patients had a sigmoid sinus wall or jugular bulb anomaly too.
- If imaging was negative and the whooshing continues, the 2024 review's point about digital subtraction angiography is the next question to ask, not a reason to accept that nothing was found.
- If a dAVF has been found, the question that determines urgency is whether there is cortical venous drainage. Ask it directly; it is the distinction the literature is built around.
- If you have been offered embolisation, the reported outcome in this series was complete resolution of the sound in four out of five.
- If your pulsatile tinnitus comes with a new severe headache, vision loss, weakness, seizure or any sudden neurological change, that is a same-day matter and not a question for an article.