Cerebral arterio-venous malformations


A cerebral arterio-venous malformation (AVM) usually consists of a group of abnormal vessels (the " nidus ") formed around one or more shunts between the arteries and veins in the brain. The absence of capillaries at the interface between the shunts decreases vascular resistance and leads to haemodynamic perturbations at various levels : acceleration of the arterial blood flow and increase in blood flow rate,  drop in pressure in the small cerebral arteries and increased venous pressure downstream from the malformation. An AVM can cause intracerebral bleeding.


Two current studies, the " New York Islands AVM Study " and the " Scottish Intracranial Vascular Malformation Study ", have made it possible to assess the detection rate of AVM as approximately 1.5 cases per 100,000 people per year.

The incidence of AVM rupture i.e. the number of patients with a malformation discovered as a result of a first intracranial bleed, is estimated at approximately 0.55 /100,000 people per year. Extrapolated to the population of France (without its overseas territories and "departments"), the results would give an estimate of approximately 900 patients per year diagnosed with AVM. The diagnosis will be made in some 300 cases every year after the patient has suffered a first intracranial bleed.

Brain AVMs are as common in men as in women and are detected without any particular ethnic preponderance.

Clinical description

Cerebral AVMs are revealed by epileptic seizures in approximately 2 out of 3 cases. Epilepsy first occurs between the ages of 20 and 40 in most patients. An AVM revealed by a fit after the age of 60 is very rare.   In most cases, the seizures are either simple or complex focal , with or without any secondary generalisation. The onset of epileptic fits is thought to be associated with certain morphological features of the vascular malformation (location in the cortex, the arterial borderzone area, in larger malformations, and those with superficial venous drainage).

A malformation can be complicated by intracranial bleeding into the brain, into the ventricular system or sub-arachnoid space (meningal).  Intracerebral bleeds can produce transient focal neurological signs, permanent signs (hemiplegia, speech difficulties, problems with eyesight etc.), epileptic seizures, sudden unusual headaches or loss of consciousness. These symptoms vary depending on the location of the lesion and the extent of the haemorrhage. An AVM rupture is thought to occur depending on certain morphological characteristics of the malformation  such as venous drainage into th deep venous system o th brain .  

Ten per cent of AVMs are discovered after diagnostic tests for epileptic seizures or headaches, in the absence of any cerebral bleed. The frequency of the primary headaches (migraine, tension headache etc.) is, however, no different whether or not an AVM  is present. In a few cases, a link between the patient's headaches and an AVM seems to be plausible, mainly because of the very localised, unilateral nature of the headache and the absence of any other cause.

Fewer than ten per cent of patients are found to have an AMV during gradual neurological impairment in the absence of any haemorrhage.  Women are thought to be more often affected in this way and risk seems to increase with age.  The mechanisms behind this type of worsening of symptoms remain unclear.

Some ten per cent of AVMs are discovered by chance on cerebral imaging, in the absence of any symptom.


AVMs expose patients to a high risk of cerebral haemorrhage. Generally speaking, the rupture rate of AVMs is rather low. However, the risk of rupture is much higher after the occurence of an initial cerebral haemorrhage. It is thought to be approximately 1% per year for AVMs that have   never bled before, and 5% or higher for ruptured AVMs. After the first rupture, the risk of recurrence of haemorrhage may be as high as 18% during the first 12 months. Apart from the occurence of a haemorrhage, the risk of rupture is thought to be particularly linked to the presence of exclusive deep venous drainage and the location of the nidus deep within the brain.


Today, an AVM diagnosis is more often easily done using the results of cerebral magnetic resonance imaging (MRI) and magnetic resonance angiography (MRA), especially when the malformations have a diameter of more than  5- 10 mm.

An arteriography (intra-arterial injection of a contrast product to display the vessels in the brain) nevetheless remains the benchmark method for, confirming the diagnosis and providing a detailed morphological definition of an AVM. The appearance of an early vein (injected much too early during the angiography) demonstrates the direct communication between an artery and a vein within the malformation.

After a recent occurrence of cerebral haemorrhage, it may be necessary to repeat the cerebral arteriography at a later date to look for a small vascular malformation because a haemorrhagic collection may conceal a small AVM and make it invisible during the initial examination.

A functional MRI that maps the major functional areas within the brain (e.g. motor and, language function) is sometimes used to assess the risk linked to a surgical procedure.


When caring for an AVM, a pluridisciplinary approach appears to be preferable.

Treatment for epilepsy and headaches, and the monitoring of the patient's neurological condition, should preferably be prodvided by a neurologist specialising in neurovascular disease.

Curative treatments, the choice of techniques and the approaches used should be assessed by vascular neurosurgeons and neuroradiologists, sometimes with a specialised radiotherapist.

 Specific treatment strategies for arteriovenours malformations are based on three different treatment types, either alone or in any combination : embolisation of the AVM by endovascular catheterisation, surgical excision or stereotactic radiotherapy (sometimes called radiosurgery). The aim of these invasive treatments is to eliminate the risk of future haemorrhage by obliteration of the AVM vessels.

The therapeutic choice is based on the patient's clinical condition and history (ruptured or non-ruptured AVM, age, neurological examination, other associated disease etc.), and on the morphology of the lesion.

The final treatment decision is mainly based on an estimate of the spontaneous haemorrhagic risk without treatment as copared to the risk of invasive therapy. Generally speaking, when the AVM is considered to be at high risk of rupture, completee treatment of the AVM is preferable, in order to eliminate the risk of haemorrhage. When the risk of rupture is estimated to be rather low, simple neurological monitoring may be sufficient, without any specific intervention. For non-ruptured AVMs, the therapeutic decision is often difficult to reach because the benefit of invasive treatment remains uncertain. Because of this, the international, multi-centre ARUBA trial ("A Randomized Trial of Unruptured Brain AVMs") which is currently underway, aims to compare the potential risk of invasive treatment the spontaneous risk (with no invasive treatment). The international trial began in 2007 and will continue for 5-10 years until final resultats may be expected (for further information, log onto: www.arubastudy.org).