Orbital vascular lesions are a complex and varied group of lesions requiring a multidisciplinary approach to management. The historical terminology and nomenclature of these lesions is confused and often misleading. A clear, pathological-anatomical basis to classification, as proposed by the International Society for the Study of Vascular Anomalies, provides a clear standardised lexicon allowing clear communication across medical specialties and evidence-based approaches to therapy. We discuss the important primary distinction between proliferative vascular tumours and non-proliferative vascular malformations. We then focus on the group vascular malformations: the low flow capillary malformation (CM), venous malformation (VM) and lymphatic malformation (LM) and the high flow arteriovenous malformation (AVM). We relate anatomical and pathological basis to the varying clinical presentations, natural history and treatment strategies of these lesions. We look at examples of cases and examine the role of multimodality imaging in diagnosis and of interventional neuroradiology in management. Finally, we look at conditions which can mimic orbital vascular lesions and the critical importance of recognising these.
BookmarkOrbital vascular lesions: role of interventional radiology
Presenter: Fergus Robertson
I’m delighted to be here today to talk about a specific interest of mine. I have no disclosures on this subject.
Orbital vascular lesions are fascinating and have confused doctors for many years. When we look back at early publications from Victorian Britain, we find beautiful descriptions of these conditions but doctors failing to understand the underlying pathology and struggling to effectively treat these lesions as a consequence. We see emerging a wide range of descriptive terms from orbital aneurysm, orbital varix to lid haemangiomas. The legacy of this descriptive naming persists today with a profusion of non-standardised terminology across the various specialities which encounter these lesions. Terms which are often confusing and misleading, hampering communicating between the different groups. The International Society for the Study of Vascular Anomalies, a cross-specialty collaborative formed in the late 90s, have worked to standardise the lexicon. The first part of the talk will focus on the modern systematic pathology-based approach to these lesions and leading to a logical consistent terminology.
If you remember one thing from this talk it should be that vascular lesions fall broadly into two categories.
The first group are the vascular tumours, which are proliferative lesions of vascular cell origin and which we will only touch on briefly today. In paediatric practice, the commonest encountered are infantile haemangiomas which are often regressive lesions and may not require treatment. In adults there are rarer often pre-malignant or malignant conditions generally relatively easy to discriminate on imaging as solid hyper vascular lesions.
The second group, the vascular malformations, are non-proliferative malformed vascular spaces and. These vascular malformations are often difficult to manage and require a multidisciplinary approach and are the focus of the rest of this talk.
We need to return to basic principles to understand these lesions, and this is where our Victorian colleagues were struggling. If we remind ourselves of the anatomy vascular tree; a branching larger arteries, smaller arterioles and tiny capillaries with a mirrored venous system of small venous radicals collecting into larger venules and veins. Superimposed on this is the lymphatic system. An error in any part of this system can lead to a vascular malformation. These can occur anywhere in the body, but we’ll be focusing on those lesions occurring in and around the orbit.
The first category to consider is the capillary malformations (CM), where the abnormality occurs within the capillary bed. These are relatively common, low flow lesions which can occur in isolation or as part of a syndrome. We will be familiar with the cutaneous capillary malformations, such as the ‘port-wine’ stain which may be associated with Sturge-Weber Syndrome. Another large group is the capillary telangiectasias which can be associated with Hereditary Haemorrhagic Telangiectasia. The treatments for these skin lesions are mainly cosmetic and they don’t really figure in my interventional practice.
The next major category, the venous malformations (VM) do keep us busy in interventional radiology, often as part of a multidisciplinary approach. These abnormalities of venous system can be either plexiform VM within the small veins, or form large variceal cavities called cavernous VM. The term cavernous haemangioma, which is often applied to these lesions, is misleading as the suffix -oma implies a proliferative lesion, a tumour, which this clearly is not. Venous malformation is a therefore a more accurate term.
Venous malformations can present at any age but are especially prevalent in middle-aged females. They are low flow lesions, often quite inconspicuous at rest, becoming more prominent with increased venous pressure i.e. patient performs a Valsalva manoeuvre (bears down). These lesions present considerable challenges to traditional surgical techniques and many patient we encounter have failed previous surgery. Imaging appearances are quite typical, lesions in the orbit, often with associated flecks of calcification (phleboliths) on CT imaging, low flow on doppler ultrasound, distention with Valsalva. There are fairly typical MR imaging appearances of cavities with enhancing walls.
The lymphatic malformations (LM) are abnormal dilatations of the lymphatic system, which also form as plexiform networks (microcystic) or as large cavities (macrocystic) or a combination of the two. More commonly presenting in childhood these tend be larger and more conspicuous than VMs and typically are not distensible with Valsalva as they do not connect to the venous system. They typically enlarge during any illness which activates the lymphatic system (e.g. viral infection) or with episodes of intralesional haemorrhage, which can cause a dramatic clinical pictures. Macrocystic LMs appear as large, non-distensible cystic cavities on imaging, whereas microcystic LMs are more solid with lots of tiny cyst-like cavities and can be quite difficult to discriminate from tumours in certain cases.
LMs present major surgical challenges to total resection and recurrences are common. More recently sclerotherapy has emerged as an effective therapy. The abnormal vascular spaces in the lesion are punctured with ultrasound guidance using a percutaneous needle. Contrast is injected to delineate the lesion, then a sclerosing agents is injected causing cavity collapse and inflammation and fibrosis in the walls for prevent re-expansion.
Sclerotherapy requires caution and is not without risks. The orbit is a confined bony space and any inflammatory process will potentially cause swelling, leading to critical ischaemia of delicate orbital structures. For this reason, a less inflammatory sclerosing agent, bleomycin is preferred n the orbit because it causes less swelling. Steroids are used to reduce the inflammation and in very rare circumstances when the pressure does rise significantly a decompressive lateral canthotomy may be required. Sclerotherapy can also cause skin necrosis and cranial nerve injuries. In VMs of the head and neck we must be aware of potential deep venous connections from the main malformation, to prevent the sclerosing agent to be travelling into the cavernous sinus and harming the important cerebral veins. However, with rigorous anatomical approach and judicious use of the less inflammatory agents, these complications are extremely uncommon.
So, we have covered the low flow lesions (capillary, venous and lymphatic malformations) and their broad management strategies.
Now we consider the high flow lesions, the Arteriovenous Malformations (AVMs) which are composed of abnormal connections between arteries veins bypassing the normal capillary network. These can be either be nidal (composed of a network of small abnormal vessels called the nidus) or fistulous (a direct connection between a larger artery and the vein) or a combination of the two. These can present at all ages, tend to be progress over time and can accelerate at times of hormonal change (typically puberty or pregnancy). I show a recent patient who’s as a young girl had a very subtle asymmetry to the lower lid, which becomes more apparent as she moves through puberty and proliferates rapidly during pregnancy into a large a problematic lesion. The lesions can be very obvious when superficial on the lids but less conspicuous when deeper within the orbit, presenting with proptosis (protruding eyeball), chemosis (red eye) and visual loss. Larger lesions which extending beyond the orbit can be even more challenging and management is often aimed at control rather than cure.
In AVMs particularly, we need very clear understanding of their anatomy and angiography is essential. This magnetic resonance angiogram (MRA) of a lower lid AVM allows us to appreciate the potential risks of endovascular treatment, but catheter angiography is essential in planning treatment. The lateral view of the internal carotid artery injection shows a very enlarged ophthalmic artery supplying the nidus. The external carotid artery injection shows additional feeding arteries from the superficial temporal, maxillary and facial arteries.
Surgical treatment of orbital AVMs is possible but intra-operative bleeding can be very difficult to control and increasingly we’re offering combined procedure. We will initially block as many of the abnormal blood vessels by endovascular or percutaneous injection of liquid embolic material prior to delayed surgical resection of the lesi0n.
I show an example of what can be achieved with trans-arterial embolisation. Here is a catheter passing from the carotid artery, through the ophthalmic artery into a branch feeding the AVM. From here an injection of embolic liquid, in this case histoacryl glue, with excellent filling of the nidus in this location allowing subsequent complete resection of the lesion 2 weeks later, once the swelling has settled.
Sometimes the arteries are inaccessible to our catheters, but we can use a percutaneous approach. In this case, the lower lid nidus is punctured directly with a needle and a newer embolic agent (PHIL™, Microvention, Ca) injected to fill the large nidus of abnormal blood vessels from a single position. These pictures from David Verity’s surgery show the embolic material filling the lesion and the lack of bleeding which aids safe and complete resection. David will tell us more about his part of the treatment in the next presentation.
To recapitulate our new lexicon, we can describe all vascular malformations using four terms: CM, VM, LM and AVM. We can discriminate these into low flow (CM,VM,LM) and high flow (AVM) based on clinical features and doppler ultrasound. Rational treatments can only be planned based on a full appreciated of anatomy, underlying pathology, natural history and potential risks of therapeutic options.
Some important conditions can mimic orbital vascular malformation and it is vital to have an experienced neuroradiologist involved in all of these cases to identify these mimics. Sometimes the lesion doesn’t lie within the orbit at all and we can be dealing with something different.
These images show a carotico-cavernous fistula, a relatively common neurosurgical problem in which the AVM lies inside the skull at the level of the cavernous sinus. This angiographic image shows direct passage of contrast from the internal carotid artery into the veins of the cavernous sinus (arteriovenous shunting), which is then congesting the superior and inferior ophthalmic veins. This will cause a very similar clinical presentation to intra-orbital AVM : proptosis, chemosis and visual impairment, but the anatomy, pathophysiology and treatment strategy are very different.
In other cases of suspected orbital vascular malformation there be no AVM at all. This nine-year-old boy was referred to me with distensible left proptosis. The orbit was very full and pulsatile and became more prominent when he cried. The MR angiogram showed prominent orbital veins but no enlarged arteries so perhaps this is a venous malformation. A catheter angiogram was performed to clarify. This lateral view of the carotid artery injection on that side shows a normal capillary phase i.e. there was no arterio-venous shunting. If we follow through the phases of the angiogram it’s only when we get to the normal venous drainage of the brain, that we see that almost all of the venous blood from that hemisphere is draining via the cavernous sinus and out to the phase through the ophthalmic veins. We call this a developmental venous anomaly, meaning that this supplementary venous drainage pathway has taken over a dominant role draining the brain, probably due to interruption in the normal venous drainage pathways into the jugular veins at the base of the skull. Any attempt to interfere with this anomalous orbital venous drainage, is likely to have catastrophic implications on the underlying brain, so the treatment is conservative here. We need to know what we are dealing with here; we need to image understand these lesions fully.
The three key points to take away from this talk,
Standard terminology: to work effectively as team, we need a common language so that we all know what we’re talking about and clear consistent terminology based on a deep understanding of fundamental pathology and anatomy.
When planning managment we need a clear goal of treatment and a full understanding or potential hazards are as we’ve seen some worrying hazards and we need to be aware of those.
It’s a team effort: these are complex lesions an no one team has the skillset to treat all of these patients, and it really is a team effort. Here are the key specialities I believe are required to support a successful service.
Q & As section
Question 1
Do you normally do lung function test in young children? And if you do sometimes a difficult to do the lung function test for these children? Do you have a lifetime dose for bleomycin that you adhere to?’
Answer
You refer to the potential dose- dependent respiratory side effects of Bleomycin shown from cancer chemotherapy trials. Historically we have attempted lung function tests where possible before and after treatments. Now, many centres including ours are moving away from pulmonary as the doses used in sclerotherapy are much lower than in oncology and we have seen no ill effects at the doses we use. As we don’t reach oncological doses, we don’t work to a life-time dose limit. We have to have a written bleomycin protocol – an important we avoid tape and sticking-plaster on the patient’s skin to avoid skin rashes.
Question 2
Can I ask with brain AVMs, we fear future brain haemorrhage, do any of these orbital vascular anomalies bleed at all?
Answer
Certainly, orbital vascular malformations can present with intralesional bleeding or thrombosis particularly if there are large vascular spaces. This can cause mass effect and compression of orbital structure – a potential surgical emergency. The high flow lesions can bleed from ulcerated skin or mucosa or from superficial varices and the degree of blood loss can be significant if unchecked. Brain AVMs are dangerous in a different way as in those bleeding into the brain causes raised intracranial pressure which can be life threatening.
Question 3
There are no lymphatics in the orbit. Where are we on this one?
Answer
If you look histologically, there are some lymphatic vessels in these anomalous orbits. These are not normal orbits. The normal orbit doesn’t really have lymphatic markers, but these orbits do have some of the immunohistochemistry of lymphatic like-vessels.
Question 4
One comment about the Onyx that you described; because my experience with lesions that have been previously embolised with Onyx is you get incredible sparking when you use a diathermy to excise them in a very small space like the orbit it could be really problematic. I’m not sure Onyx is necessarily the best agent.
Answer
Thanks for raising that. The issue of Onyx sparking with the use of diathermy has been shown to relate to tantalum, its radiographic opacifier and that what causes the sparking. This is seen less with lower power and with bipolar diathermy rather than monopolar, so those can be modified. Fortunately, this issue is less relevant as Onyx has largely been superseded in orbital treatments by PHIL embolic agent. PHIL uses iodine as its radiographic opacifier and there is no sparking with diathermy. The other advantage is that PHIL is a pale cream colour and gives a much better cosmetic result in superficial lesions compared with jet black Onyx.
Fergus Robertson
Bio: Dr Fergus Robertson is a consultant interventional and diagnostic neuroradiologist working at the National Hospital for Neurology and Neurosurgery and Great Ormond Street Hospital (GOSH) for Children in London UK. His specialist interests include management of vascular disorders of the brain, orbit and spine, neuro-oncology and non-vascular spinal disease.He is one of two dedicated UK paediatric interventional neuroradiologists and supports several nationally commissioned highly specialised services at GOSH including the national vein of Galen malformation service, management of childhood cerebral AVM and retinoblastoma therapy.
More videos in this series
About SOA
The St John Ophthalmic Association was established in 2015 and is formed of clinicians, nurses, managers, and IT and healthcare professionals from across the world. The group exists to coordinate the post-graduate ophthalmic activities of the Order of St John, uniting supporters from all professional walks of life in support of The Order of St John.
The SOA not only supports the strategy of the Hospital Board in advancing teaching, training and research across all branches of the St John Eye Hospital Group (SJEHG), but also supports the developing St John Ophthalmic Hospital in Soweto, Johannesburg.
Bookmark