Arteriovenous Malformation (AVM) Treatment
Arteriovenous Malformation (AVM) Treatment
An AVM can be treated in three ways, and the three are not versions of each other. Microsurgical resection removes the malformation at open surgery and eliminates it the moment the operation ends. Endovascular embolisation closes part of it from inside the vessel, and is more often a stage before something else than a treatment that stands alone. Stereotactic radiosurgery delivers focused radiation that closes the vessels slowly — over one to three years — and is used chiefly for small malformations in places surgery cannot reach safely. Choosing between them is not a matter of preference. It is a matter of what the anatomy allows, and of how much risk the patient would rather carry: the risk of an operation now, or the risk of a malformation that is still there for years.
The first decision is whether to treat at all
This is the question that separates an AVM from most other structural lesions of the brain, and it has to be asked before technique. In the randomised trial of unruptured brain malformations — ARUBA — patients assigned to conservative management had better outcomes over the period studied than those assigned to intervention, and the trial was stopped early on that basis. That is a real finding and it should be given to patients rather than argued away.
It has boundaries, and they are worth stating precisely. The trial enrolled unruptured malformations, so it does not tell us that a malformation which has already bled should be left alone. The follow-up period was short relative to a lifetime of risk, and it was shorter than the latency of radiosurgery itself — a technique whose benefit cannot appear inside the window the trial measured. The intervention arm was heterogeneous, and long-term extension follow-up has been reported since. Taken together, the finding is strong enough to make observation a legitimate recommendation and not strong enough to make treatment a mistake.
In practice the decision is made on the lesion and on the patient rather than on the trial alone. Favouring treatment: a low grade, a previous bleed, a progressive deficit, a young patient with decades of accumulated risk, and an associated aneurysm. Favouring observation: a high grade, a deep or eloquent location, an older patient, and a malformation found by chance in someone who has never had a symptom from it. What should not happen is a decision made on the diagnosis alone, or one in which the alternative of watching was never explained.
Microsurgical resection: removing it, now
Resection is the oldest treatment and the only one that is immediate. Under general anaesthesia the skull is opened over the malformation, the feeding arteries are identified and closed, the nidus is removed, and the draining vein is divided last. When the operation is finished, the malformation is gone, and an angiogram is performed to confirm it.
Its advantage is that it removes the risk on the day. For a low-grade malformation — small, superficial, away from eloquent cortex — the risk of the operation is lower than the lifetime risk of leaving it, and surgery is the treatment most specialists would recommend. The disadvantages are the ones any craniotomy carries: an incision, a hospital stay, a recovery measured in weeks, and the possibility of a deficit from operating near or within brain that matters.
The grading scale exists largely to make that last risk predictable. It was built to estimate what removal would cost, and it is the reason a specialist can say, of a grade I or II malformation, that surgery is curative with a low chance of a new deficit — and, of a grade IV or V one, that the same operation would carry a risk most patients would decline.
Endovascular embolisation: from inside the vessel
A catheter is guided to the feeding arteries and the vessels supplying the nidus are occluded with a liquid embolic agent or with coils. It is done through an artery in the groin or the wrist, under general anaesthesia, and it leaves no incision.
What it is used for is frequently misunderstood. Embolisation alone is curative in a minority of malformations — generally small ones with a single, accessible feeding artery — and in most cases it is a stage rather than a treatment. Before surgery it reduces flow and closes the deep perforating feeders that are hardest to control, making the operation safer and shorter. Before radiosurgery it reduces the volume of the nidus that has to be irradiated, which raises the chance of obliteration.
Its risk profile is its own. Vessels that supply normal brain run near the nidus, and closing them causes a stroke; the draining vein can occlude before the nidus is closed, which raises the pressure inside it and can precipitate bleeding. Partial embolisation that is not followed by another treatment leaves a malformation that has been altered rather than removed, and the evidence does not support treating an AVM part-way and stopping.
Stereotactic radiosurgery: closing it over years
Focused radiation — delivered as a single session on a dedicated unit — damages the lining of the abnormal vessels so that they thicken and close over time. There is no incision and no admission, and the malformation is not touched.
The cost is time, and it is the thing patients must be told before consenting. Obliteration takes one to three years. During that interval the vessels are closing but they are not closed, and the risk of bleeding during the latency period is not removed — it is the same risk the patient had before treatment. Radiosurgery is therefore a treatment that requires the patient to wait, and the wait is not a neutral one.
Its place is specific and genuine: small malformations, generally under about 3 cm, in locations where surgery would cost a function the patient cannot spare — deep in eloquent brain, in the brainstem, or in the thalamus and basal ganglia. Reported obliteration rates for well-selected lesions of that size are commonly quoted in the range of 70 to 80 per cent at three years, and lesions that are not obliterated are usually re-treated rather than abandoned.
Why most treated AVMs are treated more than once
The three techniques are often described as alternatives, and in a referral letter they are often offered as if the patient should pick one. In practice, a malformation complex enough to need referring is often complex enough to need two of them, staged over months.
The common shape is embolisation first and then something definitive: reduce the flow and close the awkward feeders, then resect or irradiate what remains. A second common shape is radiosurgery followed, years later, by embolisation or surgery for a residual nidus that did not close. And a third is surgery for the haematoma at presentation, followed by definitive treatment of the malformation once the patient has recovered from the bleed.
What this means for a patient planning treatment is that the honest unit of planning is not a procedure but a sequence, with a period of months in the middle and imaging along the way. It is also the reason a single-technique quote, given without the plan around it, is worth interrogating.
After a bleed: two problems, two timetables
When a malformation has bled there are two separate things to deal with, and they are not dealt with at the same time. The first is the haematoma and its consequences: pressure inside the skull, blood in the ventricles, hydrocephalus, and the deficit the bleed has caused. The second is the malformation itself.
In the emergency phase the haematoma takes priority. If it is large enough to raise pressure or to shift the brain, it is evacuated, and a drain may be placed for hydrocephalus. Blood pressure is controlled. If an associated aneurysm is present and is thought to be the source, it tends to be secured early, because it is the lesion most likely to bleed again soon.
The malformation is usually treated later, once the patient has stabilised and the anatomy has been mapped. That interval is not neglect — operating into an acute haematoma in swollen brain carries a higher risk than operating weeks later — and it is used to obtain the angiography and the functional mapping that the plan depends on. Where the malformation is treated during the same admission, it is usually because the bleed was small and the patient is well.
Recovery, and what improvement actually looks like
After resection, recovery from the operation itself is measured in weeks; after radiosurgery there is no recovery from the procedure, only a wait of years. After a bleed, the shape is altogether longer, and it is the one families most often underestimate.
A deficit from a bleed improves over months, and improvement is usually largest early and continues well beyond the first months. The mechanism is not the repair of dead tissue but the recruitment of other areas and the resolution of the swelling and dysfunction around the haematoma. Most of the eventual functional outcome is built in rehabilitation rather than in the first week, which is why an early review is a checkpoint rather than a verdict.
Specific deficits have specific trajectories. Weakness often improves substantially; language recoveries are slower and the more variable; and visual field loss tends to be the most persistent. Seizures, when they develop after a bleed, are usually controlled with medication. What patients should be told is that the trajectory is upwards and slow, that it is measured in months, and that stopping rehabilitation early is the most common avoidable reason for a worse result.