Brain AVM Surgery - Arteriovenous Malformation Surgery

An unruptured brain AVM is the rare lesion where leaving it alone is a serious medical option.

Surgery ends the bleeding risk on the day it succeeds, and it charges the whole price of that on the same day. Deciding between those two shapes of risk is what this page is for.

What a brain AVM is, and why it bleeds

A brain arteriovenous malformation is a plumbing fault you were born with. Arteries carry blood at high pressure, veins are built to receive it gently, and a capillary bed normally sits between the two to step that pressure down over a few millimeters of tissue. In an AVM the step is missing. Feeding arteries empty straight into draining veins through a tangle of abnormal vessels called the nidus, and those veins take arterial pressure they were never built to hold.

That is the entire problem.

Everything else follows from it. Vessel walls under the wrong pressure stretch, thin and eventually tear, which is how an AVM bleeds. Blood racing through the shunt is blood not reaching the brain around it, which is why headaches and slowed thinking show up in people whose lesion has never ruptured. Irritated cortex at the rim of the nidus fires abnormally, which is how a third of patients arrive at a neurologist with a first seizure and no idea anything was wrong. And because the lesion has been sitting there since before you could walk, the brain has often rewired around it, so an AVM parked in what a textbook labels speech territory does not always behave the way the textbook predicts. Three doors lead to the diagnosis. A hemorrhage. A first seizure. Or a scan ordered for a headache, a head injury or something else entirely, which is the group this page spends most of its time on.

The risk of leaving it alone

The honest starting number comes from a meta-analysis that pooled nine studies, 3,923 patients and 18,423 patient years of follow up. Across everyone in it, an untreated brain AVM bled at 3.0 percent a year. Split by history the picture separates sharply, because a lesion that has never bled ran at 2.2 percent a year while one that had already bled ran at 4.5 percent. Those two figures do most of the work in every conversation that follows, since they set the thing an operation is being measured against. A longer view makes the same point in a form people find easier to hold in their heads. Add the yearly risk up and an unruptured AVM carries something near a 16 percent chance of bleeding across ten years and 29 percent across twenty, while a lesion that has already ruptured sits closer to 35 and 45 percent over the same two spans. When a bleed does happen, around 42 percent of those events leave a new permanent deficit or kill the patient outright.

Read that last sentence twice. It is the reason anyone operates at all.

Now read it against the other side of the ledger. A craniotomy on a healthy 30 year old who has never had a symptom carries its own chance of leaving that person permanently worse, and that chance arrives on one Tuesday morning instead of spreading thinly across the next forty years.

What moves your yearly risk

Averages describe populations and your angiogram describes you. The same pooled analysis pulled apart which features actually shift the yearly number, and four of them did.

The four that count

Having bled already multiplies the yearly risk by around 3.2, the largest single factor in the dataset and the reason a ruptured AVM gets treated far more aggressively than an identical lesion that has stayed quiet. Deep location, meaning the basal ganglia, the thalamus or the brainstem, roughly doubles it at a hazard ratio near 2.4. Drainage running exclusively through deep veins does much the same, again around 2.4. An aneurysm sitting on a feeding artery lifts the risk by about 1.8, and those aneurysms are also the part that tends to rupture first.

Two features people expect to matter did not.

Small size showed no significant association with hemorrhage in that analysis, despite a long standing belief in the opposite direction. Older age carried none either. A patient told their AVM is small and therefore safe has been handed a statement the pooled evidence does not support, and one worth putting back to the person who said it.

The trial that split the field

In 2014 the Lancet published a study neurosurgery is still arguing over a decade later. ARUBA randomized 223 adults who had an unruptured brain AVM either to medical management alone or to interventional treatment, which meant surgery, embolization, radiosurgery or some stacked combination of the three.

Recruitment was halted early. After a mean follow up of 33.3 months, death or symptomatic stroke had reached 11 patients on medical management against 35 on intervention, a hazard ratio of 0.27 with a confidence interval running from 0.14 to 0.54. Put as percentages that is 10.1 against 30.7. Strokes ran 12 against 45, and non stroke neurological deficits 1 against 14. The monitoring board stopped randomization for superiority of doing nothing, a sentence that does not turn up often in surgical literature.

Neurosurgery took that badly, and rightly so.

A generation of practice had assumed that an AVM found by accident should come out, on the logic that yearly bleeding risk compounds across a lifetime and a young patient has a great many years to compound it in. ARUBA answered that the compounding was being outrun by the treatment itself.

Where ARUBA stops describing you

Trials answer the question they were built to answer and nothing beyond it. Four features of ARUBA decide how far its result reaches toward the person reading this page, and each is worth taking on its own.

What the surgical literature answered back

A 2017 critical review set the operative results against the same natural history figures. For a lesion with no deep venous drainage sitting in non eloquent brain, the eight year risk of an unfavorable surgical outcome ran from about 1 percent at 1 cm up to 9 percent at 6 cm. Add either deep drainage or an eloquent location and that same range became 4 to 35 percent. Add both and a lesion of only 1 to 3 cm sat somewhere between 12 and 38 percent. The conclusion drawn from it was narrow and specific rather than triumphant. A patient with a low grade AVM who can reasonably expect eight more good years does better with surgery in an experienced center than left untreated, and the qualifiers in that sentence are doing real work.

So what actually settles it

Your age, your grade, and your appetite for a risk you choose against a risk you merely carry. No web page settles that, this one included.

What a page can do is tell you which four numbers to bring into the conversation. The diameter of the nidus in centimeters, whether the draining veins run deep or superficial, whether the brain immediately around it is eloquent, and whether it has ever bled. All four sit in your radiology report already, waiting to be read out loud.

Reading a Spetzler-Martin grade

In 1986 two surgeons proposed a scale that has outlived nearly every competitor, and your file almost certainly carries a number from it. The Spetzler-Martin grade adds three points together. Size of the nidus, whether the venous drainage is deep, and whether the brain next to it is eloquent.

Size scores one point under 3 cm, two points from 3 to 6 cm, and three points above 6 cm. Deep venous drainage adds a point. An eloquent location adds another, and eloquent here has a defined list behind it, covering sensorimotor and language and visual cortex, the hypothalamus and thalamus, the internal capsule, the brainstem, the cerebellar peduncles and the deep cerebellar nuclei. The total runs from one through five, with a sixth category held back for lesions judged inoperable.

What a grade is and is not

A grade predicts surgical difficulty. It does not predict bleeding.

That distinction gets lost constantly, in clinics as well as online. Two patients holding the same grade III card can face completely different operations, since a 5 cm superficial lesion in quiet cortex and a 2 cm one buried in the thalamus both add up to three. The number works as shorthand a surgeon uses to sort cases and plan a day. It carries no verdict on whether you should have surgery in the first place, and it was never designed to.

Surgical risk by grade and size

Grades I and II are where microsurgery is at its strongest. Complete removal is the expectation rather than the hope, the cure becomes permanent the moment the last feeding artery is clipped and the angiogram confirms it, and the published risk of a lasting deficit stays low enough that the natural history overtakes it within a few years for anyone young. Grade III is the argued middle, and the argument turns on which kind of grade III you have. A small deep one behaves nothing like a large superficial one, and a careful surgeon separates the two before offering anything at all. Grades IV and V are where enthusiasm properly collapses. The risk of a permanent deficit climbs steeply, the natural history has not budged, and a substantial body of opinion holds that these lesions are left alone unless they have bled more than once or are causing a deficit that keeps worsening. That position was reached well before ARUBA and never needed ARUBA to justify it.

Find out which of those describes yours.

What the operation involves

Removing an AVM is not tumor surgery, and the difference matters for what your consent form actually commits you to. A tumor gets hollowed out from the inside and collapsed inward. An AVM cannot be entered at all, because opening the nidus partway through an operation produces bleeding nobody in the room can control. Consent forms for the two operations therefore describe different failure modes, and the AVM version turns on losing control in the middle of the procedure instead of on how much tissue was left behind at the end.

The one rule the whole operation obeys

Feeding arteries come first, interrupted one at a time while the surgeon works around the outside of the tangle. The draining veins are left until the very end. Taking a major drainer early converts the nidus into a pressurized bag with an inflow and no way out, and it ruptures on the table. Sequence, operating time and the surgeon's apparent slowness all follow from that single ordering rule, so a resection that looks unhurried from the anesthetist's side of the drape is behaving exactly as it should.

1. Positioning and navigation. The angiogram and MRI load into a navigation system, the head is fixed in a clamp, and the opening is planned to reach the nidus while disturbing the least normal brain on the way in.
2. Exposure. A bone flap is turned over the lesion and the covering opened. Where the AVM reaches the surface the abnormal red veins announce themselves immediately, and they are deliberately not touched.
3. Circumferential dissection. The surgeon works around the nidus in a plane of scarred tissue, taking feeding arteries as each one appears. This is the long part of the day.
4. Deep feeders. Arteries entering from underneath are the difficult ones, small and fragile and reached last of all. Bleeding here is what stretches an operation past its planned finish.
5. Venous division. Once the tangle has gone soft and slack, the drainers are divided and the lesion comes out in one piece.
6. Confirmation. An angiogram on the table, or within the first day or two afterward, shows whether any nidus remains. Residual nidus still bleeds, so this step is never a formality.

Published series put the resection at four to eight hours, with grade driving that more than any other variable. A small superficial lesion finishes in three. A deep one drawing feeders from several arterial territories runs past eight and sometimes gets deliberately staged across two sittings.

Embolization before the craniotomy

Some AVMs get a catheter procedure first. A microcatheter is threaded up from the groin into the feeding arteries and a liquid embolic agent is injected to block them, which cuts the flow the surgeon has to fight and can turn an eight hour operation into a five hour one.

One to three sessions covers the usual range, spaced days apart, with the craniotomy following soon after the last of them so the lesion has no time to recruit fresh supply, and embolization on its own cures only a minority of brain AVMs, which is the honest framing patients are rarely given. Used alone it closes the nidus completely in a small fraction of cases. Used as preparation for surgery it earns its place properly.

It also carries a stroke risk of its own, and that risk is not zero.

The three routes compared

Three routes exist and they are not interchangeable. Each buys a different thing on a different timetable at a different price in risk, and the gap between those timetables is where most of the confusion starts.

Timing is the real difference

Surgery is the only route that finishes on the day it happens. Once the confirmatory angiogram comes back clean the bleeding risk is gone permanently, and nothing has to be waited out. Radiosurgery works by making the vessel walls thicken and close over two to three years, and throughout that latency the AVM goes on bleeding at close to its untreated rate. A study pooling 2,320 radiosurgery patients across eight centers measured exactly that. The hemorrhage rate fell from 15.4 per thousand person years before treatment to 11.9 after it, and once obliteration was confirmed the rate dropped to 6.0 against 22.3 for lesions still patent, which says plainly that the protection comes from closure rather than from the radiation itself. Where a nidus has failed to close by the three year mark the choices narrow to a second radiosurgery session, a delayed operation, or living on with a risk you were told would be gone by now.

This table scrolls sideways on a narrow screen. Swipe or drag to see every column.

• Which route is being recommended for my grade, and what finding would change that recommendation?
• If radiosurgery, what obliteration rate is expected at three years for a nidus this size, and what is the plan if it fails?
• If surgery, how many lesions in this location has this surgeon resected?

If your AVM has already bled

Everything above shifts once there is blood on the scan. The yearly rate doubles to 4.5 percent, the twenty year cumulative figure climbs toward 45 percent, and prior hemorrhage stands as the strongest single predictor in the pooled data at a hazard ratio of 3.2. ARUBA excluded these patients entirely. So the trial quoted at people as a reason to avoid surgery has nothing whatsoever to offer anyone whose AVM has ruptured, a point that vanishes every time the study gets compressed into a headline. Timing then becomes a question in its own right. Surgery in the first days after a large bleed is sometimes forced by the clot itself and the pressure it is creating inside a closed skull, while a patient who is stable and not deteriorating gives the team room to wait two to four weeks, letting the hematoma break down and the swollen brain settle before a planned resection.

The clot helps in one way. It has already opened a corridor the surgeon would otherwise have to make.

Recovery, week by week

Your first night belongs to intensive care, and blood pressure explains why. After a large AVM comes out, brain that has been quietly starved for decades suddenly receives full arterial flow again, and vessels accustomed to scavenging can respond to that very badly.

The shape of the next three months

Recovery from an AVM resection follows a shape most patients recognize once somebody lays it out for them, and the table below sets each stage against what it will ask of you.

This table scrolls sideways on a narrow screen. Swipe or drag to see every column.

Fatigue outlasts everything else in that table.

Patients who feel fine at three weeks get flattened by an ordinary Thursday afternoon at four, and that is a normal feature of brain surgery rather than a warning sign, so plan your return to work around that pattern instead of around how strong you feel on the morning you are discharged.

Seizures, driving and medication

Around a third of AVM patients arrive with epilepsy as the symptom that led to the diagnosis, and a proportion of those who never had a seizure beforehand will have one after surgery. Anticonvulsant medication runs as standard for a period afterward and then gets tapered, and the taper is a decision your surgeon makes rather than one you make at home.

Driving is the part people forget to raise until it is too late to plan around, and it is governed by national law rather than by your surgeon's opinion, and countries impose a fixed seizure free interval before a license comes back, six to twelve months in many of them, with the clock starting at the last seizure instead of at the operation.

Insurance and employment run on the same clock in several countries, so a job that involves driving, working at height or operating machinery can be paused for the same interval, and an employer told in the first week handles it far better than one who finds out in month seven.

Get your own country's rule in writing before you fly.

Cost and what a package covers

No figure appears in this section, and the omission is deliberate. A number written today would be wrong by the time you read it and misleading in the meantime, so what follows is the shape of a quote and the items that decide where it lands.

Grade drives the price further than anything else does. A single stage resection of a grade II lesion, an operation with two embolization sessions in front of it, and a case that has to be staged across two admissions are three different pieces of work with three different operating room bookings and three different lengths of stay behind them.

Compare quotes on those lines rather than on the headline.

Having AVM surgery in Istanbul

Patients travel to Biruni Hospital for this operation from Europe, the Gulf, North Africa and Central Asia, and the practical worries they raise turn out to be identical every time. Who explains the angiogram. Who answers a message at two in the morning when something feels wrong and the ward has gone quiet. Whether a husband or a wife can stay overnight.

Staff in the international patients office work in English, Arabic, French, Russian, Serbian, Romanian and Spanish, with interpreting arranged for other languages on request. One coordinator stays with you from your first message through to discharge and stays reachable on WhatsApp after you have flown home, which is the part that earns its keep when a local scan needs a second opinion four months later. Patient rooms carry a companion bed, so one person sleeps beside you every night of the admission. Hotel and transfers are arranged for the nights on either side of the stay. A female physician can be requested. An invitation letter for the visa goes out around ten days before you travel. Halal, vegetarian and diabetic meals are prepared as routine and a prayer room is open on site. Radiotherapy and chemotherapy services sit on the same campus, which matters for the patients whose imaging turns out to show something other than an AVM once it is reviewed properly.

Follow up after you land back home runs on a written schedule instead of on good intentions. Local scans and reports go back to the team that operated for review, a photograph of a clinic letter or a question on tapering the anticonvulsant reaches your coordinator on the number you have used from the beginning, and the imaging intervals are written into your discharge summary so your own neurologist works from the same plan as the surgeon.

Sending your angiogram, MRI and reports for a free review costs nothing and commits you to nothing. For an unruptured AVM in particular, a second opinion on whether to operate is worth considerably more than a second opinion on how.

That second opinion is what this whole page has been arguing for.

Brain AVM surgery FAQ