Hemifacial spasm: Surgical management

Surgical management of Hemifacial Spasm

I have no disclosures.

I am grateful to Mr Ludvic Zrinzo, one of my consultant colleagues at Queen Square, for many of these slides.

Moving to the lateral skull base, at least in terms of the site of the pathology, but still with its effects visible around the eye, David has asked me to talk about surgical treatment for blepharospasm; I am really going to focus on microvascular decompression for primary hemifacial spasm.

Hemifacial spasm was first described in 1875, almost certainly in relation to a giant vertebral artery aneurysm- thankfully not the domain of neurosurgeons these days, but rather the interventional neuroradiologist. The classical features of hemifacial spasm however were described by Gowers in 1886 and in the early 1900s it was observed to be worse during periods of stress. The term hemifacial spasm was coined in 1905 by Babinski.

Hemifacial spasm is more common in women and happens in about 10 per 100,000 of the population, generally in middle to late age; it is usually sporadic but very occasionally can be seen in familial cases; some really unfortunate patients may actually develop bilateral symptoms.

The pathophysiology behind the cause of hemifacial spasm is still open to discussion. There are three main hypotheses:

(1) The peripheral theory, where it is suggested there is direct transmission between demyelinated VII nerve fibres;

(2) The central theory where there is hyperexcitability of the facial motor nucleus due to an irritative feedback from peripheral lesions. This is obviously generated in the main by vascular conflict, at least in the context of primary hemifacial spasm. Neither of these theories explains perhaps why hemifacial spasm is worse during stress, and also why it will stop immediately after microvascular decompression.

(3) A third pathophysiological hypothesis gaining credence, which may account for why symptoms do get better immediately after surgery and are worse during times of stress, is the suggestion that hemifacial spasm may be related to aberrant connections between the sympathetic nerve plexus sited around the arterial vessel wall and the facial nerve fibres which they are in direct contact with. This may account for the increase in symptoms seen during stress when there is increased sympathetic activity. The vascular conflict is sited at the junction between the central and peripheral nervous systems and this is where it is felt that the nerve may be particularly sensitive to demyelination, and therefore the effects of vascular conflict.

Hemifacial spasm usually begins in the upper “orbicularis oculi” and typically with eyebrow elevation. This can be a useful way of confirming true hemifacial spasm. If you see people or even try yourself to create hemifacial spasm, it is actually very difficult to raise your eyebrow while contorting your face. It generally then spreads down the whole of the face. Very rarely, it will go the opposite direction but classically it will start at the top and spread downwards. It does persist during sleep because it is a peripheral nerve at that point, and occasionally patients will describe a “clicking” noise in their ear due to the stapedius muscle activity.

It is important to establish when seeing patients whether they have primary hemifacial spasm related to vascular conflict with compression of the VII nerve at the root exit zone (generally by a branch of  the anterior inferior cerebellar artery (AICA), vertebral artery (VA) or very rarely, the superior cerebellar artery (SCA)), or whether they have secondary hemifacial spasm caused by structural lesions, which must be excluded. These structural lesions are generally found in the cerebello-pontine angle or in the brainstem itself.

Treatment, in the vast majority of people, is reassurance. If treatment is necessary and symptoms relatively mild, Botulinum toxin injections are commonly used; however this is a temporary solution in a long-term condition which requires multiple treatments and trying to get symmetry of the face in this context is often very difficult. Alternatively, the surgical solution is a microvascular decompression (‘MVD’).

Radiology is really important in these cases. It is important in order to exclude secondary causes of hemifacial spasm as well as for patient selection and pre-operative planning. We generally use a 3T MRI heavily T2-weighted sequence. Axial slices are generally more commonly used, however both coronal and sagittal images are useful.

One of the most important things, at least in terms of improving post-operative symptomatic relief, is to make sure that the vascular conflict has been removed along the whole length of the nerve. One thing that we need to remember as surgeons, is that the facial nerve comes out of the ponto-medullary sulcus and actually runs along the surface of the pons before it transits laterally out into the cistern and into cerebrospinal fluid (CSF) where it is more easily visible.   It lies plastered to the surface of the pons in the ponto-medullary sulcus which intra-operatively is at the depths of the surgical field.   Many failures to gain symptomatic relief are likely to be due to the fact that the vascular conflict may lie in the ponto-medullary sulcus rather than perhaps the more obvious conflict seen in the cistern. It is also about ensuring that whole zone of transition between central and peripheral nerve has been decompressed.

Surgical approach is through a routine retromastoid posterior fossa approach, with the patient in a park-bench position in pins. A standard retroauricular curvilinear incision is made and a small craniectomy/craniotomy fashioned, exposing the angle between lateral sinus and the sigmoid sinus.   Any bridging veins are disconnected allowing release of CSF. The cerebellum is then able to relax away allowing good exposure without the need for cerebellar retraction. The next angle to expose is more inferiorly, exposing the lower cranial nerves. Often, particularly with a torturous vertebral artery, it is important to make sure that the full length of the vertebral artery has been mobilised allowing manipulation of the vessels and branches easily.

Perhaps one of the most important things is the use of intra-operative electrophysiology monitoring of both the VII and VIII nerve. You can spend all the time in the world looking at your images preoperatively and trying to establish where you actually think the conflict is, but when you are operating down a deep dark hole, it can be very difficult to be really clear that every point of the nerve has been decompressed. Electrophysiology however will provide confirmation one way or the other.

Classically surgeons look at lateral spread. An impulse is given into the orbicularis oculi, via the zygomatic branch, the impulse then heads back to the brainstem and back out again. There is a slightly delayed response in orbicularis oris and mentalis, the other branches that are also being monitored. This is important because it confirms the diagnosis and confirms the necessary monitoring is in place. The loss of this lateral spread helps to confirm decompression of the nerve.

An additional electrophysiology assessment that improves the chances of remission (by more accurately predicting adequate decompression of the necessary vessel), is by stimulating the vessel rather than the nerve itself. This is the Z-L response, where the vessel itself is stimulated, and which disappears in all branches if the nerve is satisfactorily decompressed. Outcomes show that the improvement in symptoms is much better and sustained longer if both of those electrophysiology markers are used to confirm adequate decompression of the nerve.

Different methods can be used to decompress the nerve. Some use Teflon to separate the nerve from the vessel. An alternative approach is to try to hold the vessel away from the nerve. This can be done using a dural sling held by an aneurysm clip, or a fenestrated aneurysm clip where the vessel sits within the fenestration and the clip is attached to dura.   Others use Tisseel to ‘stick’ the vessel to dura on a temporary basis whilst relying on Teflon to create an inflammatory response resulting in a more permanent solution.

There are obviously risks associated with this procedure: facial nerve palsies are seen in approximately 10%, but most will recover. There is a risk to lower cranial nerve function but again that will generally recover. Hearing impairment does occur, although it may be that this is not always assessed thoroughly for pre-operatively.

Three key points:

Microvascular decompression for hemifacial spasm has been reported to have more than 90% cure rates

High resolution imaging allowing excellent visualisation of the nerves and the conflicting vessels is mandatory

The increased use of neurophysiology, looking at several indices, will improve the likelihood of post-operative symptomatic relief still further

Q & As section

Question 1
With the partner condition trigeminal neuralgia, which is another neurovascular conflict problem in the CPA, you really only need a suggestion of a vessel on the scan around the trigeminal nerve to be pretty sure you’re going to get a positive result when you go in. But with hemifacial spasm, I certainly need to see a really good conflict of a vessel, not just a vessel in the general area on the scan before I think there’s going to be a positive outcome after surgery. Is that something that you’ve noticed yourself?

Answer
I think it is about really being absolutely sure that you have looked at the full length of the nerve but yes, I think there has to be a clear vessel conflict. We have to remember that we do tend to focus on the cisternal part of the nerve, but because the facial nerve has a long course on the surface of the pons itself, it is important that we look both radiologically and at the time of surgery along the full length of the nerve for the vascular conflict. I agree that it is important to see convincing evidence of a vascular conflict radiologically before accepting the patient for surgery but I also think there may be patients who are not referred because there is no obvious cisternal vascular conflict but actually perhaps when you go back and have closer look along the ponto-medullary sulcus, there is actually a vessel.

Question 2
Are you saying effectively that you need contact between the vessel and the nerve? So you really just need to create an electrical break between the vessel and the nerve, which can either be a gap or some material like your Teflon spacer. Does that not conduct or does that seem to break the circuit? It’s an electrical cross reaction process rather than a mechanical one.

Answer
There is some evidence to suggest that it may be an electrical phenomenon rather than a distortive/compressive one.   There is an anxiety of using a spacer of any material that this distorts the normal anatomy, and may increase the risks of post-operative cranial nerve palsy and/or failure to improve symptoms.

Question 3
How do you keep the vessel away permanently from the nerve? To make a gap?

Answer
People use lots of different techniques. Some surgeons will put Teflon as a physical barrier between vessel and nerve. Other people will try and hold the vessel away by using aneurysm clips.

Question 4
I haven’t come across the bipolar stimulation technique for the VII. You said that’s confirmatory. Will it change your management if you still got spread into the nerve at the end? Would you go back and explore again or how did you use that?

Answer
By using it intraoperatively, if you’re still getting lateral spread, it implies that you have not fully decompressed the nerve and I would continue to look for a vessel. By using the two electrophysiology tests, it certainly increases the chances of knowing that you have cured that patient.

Question 5
In your experience, how often do you find that it does change your management at the end? You go back and look again?

Answer
The electrophysiologist will be constantly stimulating and tells the surgeon when the lateral spread and Z-L response disappears. So the surgeon tends to continue operating until they hear that news rather than ‘going back in’ when they fail to disappear.