72u7rnal of Ncenrologj', Aeurosurgery, caitd PPsYchiatry 1 996;60:772- 77 Hemifacial spasm: a prospective long term follow up of 83 cases treated by microvascular decompression at two neurosurgical centres in the United Kingdom R D Illingworth, D G Porter, J Jakubowski Abstract Objective-To evaluate the use of microvascular decompression (MVD) for the treatment of hemifacial spasm (HFS). Methods-Eighty three patients with HFS who underwent MVD via a suboccipital craniectomy are presented. Results-Seventy two out of seventy eight patients available for follow up remained free of any spasms at a mean follow up period of eight years. Two patients con- tinued to have minor intermittent muscle twitches and three had recurrence of HFS. One patient's operation was not completed. Twenty had a transient complication and eight were left with per- manent postoperative deficits, the com- monest being unilateral sensorineural deafness. Seventy one patients declared themselves satisfied with the procedure. A causative vessel was found on the root exit zone of the seventh cranial nerve in 81 patients. Conclusion-The procedure seems to provide lasting relief for most patients. The correct operative technique is essen- tial if complications are to be avoided. ( Neurol Nreurosurg Psychiatry 1 996;60:72-77) Keywords: hemifacial spasm; microvascular decom- pression Regional Department of Neurosciences, Charing Cross Hospital, London, UK R D Illingworth D G Porter Department of Neurological Surgery, Royal Hallamshire Hospital, Sheffield, UK J Jakubowski Correspondence to: Dr R D Illingworth, Regional Department of Neurosciences, Charing Cross Hospital, London W6 8RF, UK. Received 18 April 1995 and in revised form 30 August 1995 Accepted 6 September 1995 Hemifacial spasm (HFS) is a condition con- sisting of unilateral paroxysmal involuntary tonic clonic spasms occurring in the muscles innervated by the facial nerve. The involuntary contractions usually begin in the orbicularis oculi muscle and gradually spread to the other muscles of facial expression. It is a rare condition. The average annual incidence has been reported as 0 74 per 100 000 for men and 0 81 per 100 000 for women, and the average prevalence as 7 4 per 100 000 in men and 14 5 per 100 000 in women for the white population of the United States.' Thus it is not often encountered in neurosurgical practice and results have mainly been reported by tertiary referral centres in the United States.2 Causative lesions that have been reported include aneurysms of the posterior circula- tion,5 epidermoid tumours,6 and arteriovenous malformations.7 However, most cases are due to arteries, usually the posterior inferior cere- bellar artery, compressing the facial nerve at the root exit zone.8 I' Despite the use of microvascular decom- pression (MVD) to treat HFS for many years there are still only very limited results of long term follow up. We present 83 cases of HFS treated by MVD at two neurosurgical centres by two neurosurgeons who undertake this pro- cedure only as part of a wider neurosurgical practice, and we discuss the relative merits of MVD compared with other treatments. Patients and methods Eighty three patients were treated at two neu- rosurgical centres in the United Kingdom between 1978 and 1992. Forty seven were women and 36 were men, with ages ranging from 25 to 78 years (mean 53 3 years) at the time of operation. The spasms were on the right side in 42 and the left side in 41 patients. The preoperative duration of symptoms was from eight months to 25 years with a mean duration of 6 15 years. As well as the muscular spasms, other neurological findings consistent with the diagnosis were present in 50 patients. Table 1 summarises these. After investigations including brain CT (vertebral angiography in the earlier cases) to exclude an underlying causative mass lesion, MVD was performed. The posterior fossa approach as described by Jannetta" and modi- fied by Sugita 2 was used. The patients were placed in the lateral decubitus position and a small anterior inferior craniectomy, about 3 cm in diameter, extending up to the posterior margin of the inferior part of the sigmoid sinus was performed. On opening the dura, the ninth and 10th cranial nerves were exposed as they passed from the jugular foramen towards the side of the medulla. These nerves were fol- lowed with careful microsurgical dissection back to the side of the medulla by freeing the Table I Preoperative neurological signs in 83 patienits with HFS Neurological signs \o of patien1ts Seventh nerve paresis alone 32 Seventh nerve synkinesis alone 7 Seventh nerve paresis with synkinesis 9 Hemifacial spasm with "trigeminal neuralgia" 2 Total 50 72 on March 9, 2020 by guest. Protected by copyright. http://jnnp.bmj.com/ J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.1.72 on 1 January 1996. Downloaded from
Hemifacial spasm: a prospective long term followup of 83 cases treated by microvasculardecompression at two neurosurgical centres in theUnited Kingdom
R D Illingworth, D G Porter, J Jakubowski
AbstractObjective-To evaluate the use ofmicrovascular decompression (MVD) forthe treatment of hemifacial spasm (HFS).Methods-Eighty three patients with HFSwho underwent MVD via a suboccipitalcraniectomy are presented.Results-Seventy two out of seventy eightpatients available for follow up remainedfree of any spasms at a mean follow up
period of eight years. Two patients con-
tinued to have minor intermittent muscletwitches and three had recurrence ofHFS. One patient's operation was notcompleted. Twenty had a transientcomplication and eight were left with per-
manent postoperative deficits, the com-
monest being unilateral sensorineuraldeafness. Seventy one patients declaredthemselves satisfied with the procedure.A causative vessel was found on the rootexit zone ofthe seventh cranial nerve in 81patients.Conclusion-The procedure seems toprovide lasting relief for most patients.The correct operative technique is essen-
Regional Departmentof Neurosciences,Charing CrossHospital, London, UKR D IllingworthD G PorterDepartment ofNeurological Surgery,Royal HallamshireHospital, Sheffield, UKJ JakubowskiCorrespondence to:Dr R D Illingworth,Regional Department ofNeurosciences, CharingCross Hospital, LondonW6 8RF, UK.Received 18 April 1995and in revised form30 August 1995Accepted 6 September 1995
Hemifacial spasm (HFS) is a condition con-
sisting of unilateral paroxysmal involuntarytonic clonic spasms occurring in the musclesinnervated by the facial nerve. The involuntarycontractions usually begin in the orbicularisoculi muscle and gradually spread to the othermuscles of facial expression.
It is a rare condition. The average annualincidence has been reported as 0 74 per
100 000 for men and 0 81 per 100 000 forwomen, and the average prevalence as 7 4 per100 000 in men and 14 5 per 100 000 inwomen for the white population of the UnitedStates.' Thus it is not often encountered inneurosurgical practice and results have mainlybeen reported by tertiary referral centres in theUnited States.2
Causative lesions that have been reportedinclude aneurysms of the posterior circula-tion,5 epidermoid tumours,6 and arteriovenousmalformations.7 However, most cases are due
to arteries, usually the posterior inferior cere-bellar artery, compressing the facial nerve atthe root exit zone.8 I'
Despite the use of microvascular decom-pression (MVD) to treat HFS for many yearsthere are still only very limited results of longterm follow up. We present 83 cases of HFStreated by MVD at two neurosurgical centresby two neurosurgeons who undertake this pro-cedure only as part of a wider neurosurgicalpractice, and we discuss the relative merits ofMVD compared with other treatments.
Patients and methodsEighty three patients were treated at two neu-rosurgical centres in the United Kingdombetween 1978 and 1992. Forty seven werewomen and 36 were men, with ages rangingfrom 25 to 78 years (mean 53 3 years) at thetime of operation. The spasms were on theright side in 42 and the left side in 41 patients.The preoperative duration of symptoms wasfrom eight months to 25 years with a meanduration of 6 15 years. As well as the muscularspasms, other neurological findings consistentwith the diagnosis were present in 50 patients.Table 1 summarises these.
After investigations including brain CT(vertebral angiography in the earlier cases) toexclude an underlying causative mass lesion,MVD was performed. The posterior fossaapproach as described by Jannetta" and modi-fied by Sugita2 was used. The patients wereplaced in the lateral decubitus position and asmall anterior inferior craniectomy, about 3cm in diameter, extending up to the posteriormargin of the inferior part of the sigmoid sinuswas performed. On opening the dura, theninth and 10th cranial nerves were exposed asthey passed from the jugular foramen towardsthe side of the medulla. These nerves were fol-lowed with careful microsurgical dissectionback to the side of the medulla by freeing the
Table I Preoperative neurological signs in 83 patienitswith HFS
Neurological signs \o of patien1tsSeventh nerve paresis alone 32Seventh nerve synkinesis alone 7Seventh nerve paresis with synkinesis 9Hemifacial spasm with "trigeminal
neuralgia" 2
Total 50
72
on March 9, 2020 by guest. P
rotected by copyright.http://jnnp.bm
j.com/
J Neurol N
eurosurg Psychiatry: first published as 10.1136/jnnp.60.1.72 on 1 January 1996. D
Hemtzifacial spasm: a prospective long term follow up of 83 cases treated by, microvascular decompression at two neurosurgical centres i1n the UK
nerves first from the flocculus of the cerebel-lum, and then, more medially, from thechoroid plexus of the lateral recess of thefourth ventricle. At this point the eighth nerveis found passing transversely across the line ofdissection. The seventh nerve is found there asa white elongated triangle, lying slightly caudaland medial to the eighth nerve. The compress-ing artery was usually discovered before thispoint was reached, looping across the side ofthe brainstem. Little dissection was requiredto lift the artery away from the seventh nerve,which is usually deeply grooved at the pointwhere it leaves the side of the brainstem. Thecausative artery was mobilised and held awayfrom the nerve by a piece of Ivalon sponge,lintine patty, or shredded teflon.
Intraoperative monitoring of auditory andfacial nerve functions has not been used.
All operative procedures were undertakenby RDI or JJ. Assessment of the results wasconducted by outpatient follow up andrecently by a standard questionnaire sent toeach patient. Information regarding the exactvascular relation to the facial nerve wasobtained from the operative notes. If thepatient did not respond, an attempt was madeto contact them by telephone or by an outpa-tient interview. Information on three deceasedpatients was obtained from family or generalmedical practitioners. In 78 of 83 patients fol-low up was achieved but the other five couldnot be contacted. Three are resident overseasand have not responded to letters, and twohave not been contacted in this country.
ResultsOPERATIVE FINDINGSA vascular structure was found to be com-pressing the facial nerve, which was groovedtransversely at the root exit zone, in 81(97 5%) out of 83 patients in our series. Thecompressing vessel in most cases was the pos-terior inferior cerebellar artery but mostly itwas not possible to identify the artery withconfidence (table 2). In 12 patients two sepa-rate arterial loops were causing compressionand in one patient three loops. If two or moreloops were found then all were mobilised andheld away from the facial nerve. One case hadto be abandoned because of an unstable bloodpressure due to haemorrhage from an acces-sory occipital sinus before the facial nerve wasexplored.
Table 2 Source of seventh nerve compression by avascular structure in 82 * patients with HFS
Table 3 Timningfor 10 patients who had a delay to totalcessation ofHFS after MVD
Delay to total Lenigth of tim71Xe in years ofcessation of spasm follow lup
4 days 14 34 days 7-17 days 5-57 days 8-07 days 7-02 weeks 7 73 weeks 8.83 weeks 13-02 months 16*87 months 7-3
RELIEF OF SYMPTOMSIn the 78 patients in whom long term followup was achieved, 72 (92%) were completelyrelieved of their HFS by MVD. Spasms ceasedimmediately in 62 patients and after a delay ofup to seven months in 10 patients (table 3). Inother patients spasms often stopped immedi-ately after the operation but then returned forone or two days before finally ceasing. Twopatients had major improvement but contin-ued to have minimal twitching around the eyeon the side of the previous spasms.
In response to the questionnaire, 71 of 78(91 %) of the patients declared themselves sat-isfied with the results of the operation.
Four patients had an intermittent drum-ming sound in the ear on the affected sidebefore the operation. This was relieved by theoperation in all cases.
Three cases of HFS treated by MVDrecurred and one patient's spasm remainedunchanged in the immediate postoperativeperiod. The recurrences occurred at six, nine,and 24 months. The first patient chose to haveno further treatment and the condition spon-taneously resolved. The second patientdelayed further intervention until four yearsfrom the original procedure and was re-explored uneventfully. No causative vessel wasfound. The spasm improved but recurred twoyears later. No further surgery was under-taken. The last patient had a recurrence ofHFS at two years. A further exploration wasundertaken and the facial nerve was found tohave been inadequately decompressed fromthe ectatic basilar artery at the original proce-dure. Further packing was introduced todecompress the nerve and the patient wasrelieved of spasm for two years, since when hehas been lost to follow up and thus cannot beincluded in the final results analysis.
In the patient whose HFS remainedunchanged in the immediate postoperativeperiod, a re-exploration two weeks after thefirst procedure found a small loop of the pos-terior inferior cerebellar artery at the root exitzone that had been missed at the first opera-tion. The root exit zone was decompressedwith complete relief of the spasms. Thepatient's spasm returned after two years but hedeclined further surgery.
In the patient whose operation was aban-doned before the facial nerve was exposed, thespasms continued and showed progressiveworsening.
There have been no recurrences during amean follow up period of 9 6 (range 5 5-16 8)
73
on March 9, 2020 by guest. P
rotected by copyright.http://jnnp.bm
j.com/
J Neurol N
eurosurg Psychiatry: first published as 10.1136/jnnp.60.1.72 on 1 January 1996. D
*Eighty five microvascular decompressions were performed on 82 patients.
years, in the 10 patients who had a delayedresponse to the operation. This is by contrastwith the experience of Barker et al who foundthat patients with evidence of postoperativespasm, even for short periods, had a higherincidence of recurrence.4
Five patients were lost to long term followup. One patient has not been seen since dis-charge when he was free of spasms and one
patient, who was free of HFS after MVD, was
only followed up for one month. Two furtherpatients, one seen two years and the other 18months after their operation were free ofspasms. The fifth case lost to follow up hasbeen described above.
ASSOCIATION WITH TRIGEMINAL NEURALGIA
Two patients in our series had both HFS andtrigeminal neuralgia. This association has beenpreviously described.'3 Both patients under-went MVD for each disease with resolution ofsymptoms in the immediate postoperativeperiod. Only one patient was available to fol-low up and was still asymptomatic.
FOLLOW UP
The mean time of follow up in our series iseight (range 2-17 25) years with 43 patients(54%) followed up for a minimum of eightyears.
COMPLICATIONS
There was no operative mortality in this series.Complications occurred in 28 (32 9%)
patients. There were 20 (23-5%) with a tran-sient complication and subsequent completerecovery. Eight (9-4%) had a permanent post-operative deficit. Table 4 illustrates the post-operative complications.The commonest permanent neurological
deficit after the procedure was unilateral sen-
sorineural deafness, which occurred in fourpatients. Three other patients experienced a
mild permanent reduction in hearing on theoperated side. One patient had a permanentpostoperative facial palsy. One patient hadsubjective but not objective facial numbness inthe distribution of the first and second divi-sions of the trigeminal nerve on the operatedside.
Transient facial weakness occurred in ninepatients. This was often delayed for two tothree weeks after the operation and all fully
recovered after a mean duration of twomonths (range 10 days to three months). Amild temporary hearing reduction was notedby four patients and four complained of tem-porary postoperative dizziness and vertigo.There was a single case of meningitis and oneCSF leak that spontaneously resolved. Onepatient returned six weeks after the MVD withsymptoms from a subtentorial arachnoid cyston the contralateral side. The cyst was notpresent when the preoperative scan wasreviewed. The symptoms were relieved by acystoperitoneal shunt.
In one case the procedure was abandoned,due to an unstable blood pressure from pro-fuse bleeding at the cervico-occipital junctionfrom an accessory occipital venous sinus.Exploration of the seventh nerve was not car-ried out.
DiscussionHFS is a condition consisting of unilateralparoxysmal involuntary contractions of themuscles innervated by the facial nerve. Thecondition can be very disabling with personalembarrassment resulting in self imposed socialisolation, and problems with personal interac-tion interfering with employment. In somecases the eye on the affected side may remainclosed sufficiently long to interfere with activi-ties such as driving, which require binocularvision.
Medical treatment of HFS, using carba-mazepine, clonazepam, or orphenedrine islargely disappointing.'4 16 Treatment with botu-linum A toxin for facial dyskinesiae includingHFS is well known.I7 19 This techniquerequires multiple injections into the musclesand is effective for two to three months, when itmust be repeated. Thus injections are for anindefinite time. The quoted success rate forsignificant relief of symptoms is 70-75%.'"1.The procedure can give rise to complicationsincluding ptosis, exposure keratitis, diplopia,epiphora, drooling, and strabismus. 2 Excessdosage at the time of injection can result intemporary facial paralysis. The reported com-plication rate with this technique is 2% to 14%per treatment.27'1 The method has the majoradvantage of saving patients the risk of majorintracranial surgery and is therefore often pre-ferred to MVD by both patients and physi-cians. The need to repeat the injections atfrequent intervals is however, a considerable(and expensive) disadvantage. Also, somepatients undergoing MVD after a course ofinjections have an impaired cosmetic effectdue to a loss of fine muscle movement which isoften permanent. Apfelbaum22 makes a strongpoint regarding the difference in quality of life ifa patient is not constantly reminded of hissymptoms. Although raised as an issue in themanagement of trigeminal neuralgia, this isequally applicable to the management of HFS.The only curative treatment is an operation.
Previous surgical treatment was aimed atdestructive procedures. Total or partialdestruction of the peripheral trunks orbranches of the facial nerve by surgical expo-
74
on March 9, 2020 by guest. P
rotected by copyright.http://jnnp.bm
j.com/
J Neurol N
eurosurg Psychiatry: first published as 10.1136/jnnp.60.1.72 on 1 January 1996. D
Hemifacial spasm: a prospective long ternm follow up of 83 cases treated by, microvascular decompression at two 7neurosurgical centres in the UK
Table 5 Long tern results in published series ofMVD for HFS
No in Follow up Comiplete Partial Vascular crossAuthor (ref) senres Operations (mean) relief relief No benefit Recurrence compression
Auger et at 54 NA range 3 months-i0y 44 (81%) 5 (9%) 5 (9%) 6 (11%) 53 (98%)(3 9y)
Barker et al' 612 612 range 1-20y 86%O, 5% 9% NA 612 (100%)(8y)
Iwakuma et al2 74 NA 1 month-3y 72 (97%) 1 (133%) 1 (1 3%X) 1 (1 3%) 73 (98 6%)Loeser and Chen"' 20 21 range 4 months-7y 17 (85%) 3 (15%) 5 (25%) 75%S,
(personal series) (2 5y)Loeser and Chen' 433 450 NA 880/o 5% 2% 19 (4%) NA
(literature review)Panagopoulos et alt' 29 NA 6 months-8-5y 26 (90%) NA 3 (10%) NA 29 (100%)Piatt and Wilkins 48 NA (3 5y) 30 (62 5%) 12 (25%) NA 6 (12 5%) 46 (96%)Wilkins (review)3 41 NA range 5-12y 30 (73%) 6 (15%) NA 5 (12%/,) NA
*Percentage calculations based on 78 patients for long term follow up.NA = not available.
sure or percutaneous glycerol injection substi-tutes facial weakness for abnormal facialmovement. Relief from the spasms was oftenimmediate but returned after three to sixmonths due to nerve regeneration.23 Divisionof the facial nerve beyond the stylomastoidforamen with a faciohypoglossal anastomosishas been carried out. There is a facial palsy fortwo months then recovery to reasonable sym-metry at rest. Active movement using thetongue can be learnt but mass contractionmovements can remain troublesome. In a
series of 13 patients treated by this method 12were satisfied, although the duration of followup was not specified.24
Wakasugi25 treated 239 patients by placing a
needle percutaneously into the facial nerve
trunk at the stylomastoid foramen and per-
forming a radiofrequency lesion. The facialpalsy disappeared at one to two months inmost but only 47% of patients were relieved oftheir spasms at eight months. Iwakuma et af-6noted that 30 patients who had previouslyreceived this treatment, refused further injec-tions because of the severe pain, excessiveparalysis, and early return of the spasms. Fischand Esslen favour a selective neurectomy ofthe facial nerve branches involved in thespasms.27 Using this method Iwakuma et aP6
found that HFS recurred at one year in 60%of patients and only 15% had relief of spasmfor two to three years. Ludman and Choa25described their experience with a group of 62patients, all with HFS, in whom the facialnerve was exposed within the middle ear, via a
transtympanic route. A hook was then pushedthrough the thickness of the facial nerve andmoved longitudinally. Between three and sixpunctures of the facial nerve were made in thisway. Analysis of the results showed a mean
remission of 9-3 months after the first opera-
tion but at subsequent procedures the benefitwas less and less apparent, with mean remis-sion falling to 5-6 months at the second, 2-9months at the third, and only one month atthe fourth procedure.The pathophysiology of HFS and whether
the compressing vessel is causative is debated.Jannetta maintains that the vascular cross
compression is the cause of the spasm andMoller and Jannetta29 30 and Nielsen andJannetta3l have performed preoperative, intra-
operative, and postoperative recordings inpatient cohorts to try to support this hypothe-SiS.29 31 Several other authors agree with thisexplanation.32 33 However, the concept is chal-lenged by Adams and Kaye34 and Adams,33who regard the relief of the spasm as sec-ondary to a mild degree of operative traumaand later fibrosis around the nerve.
Despite this and because of the unsatisfac-tory results from the operations describedabove, MVD as described and popularised byJannetta has become the preferred operativetreatment for HFS.
Table 5 summarises the operative findingsand postoperative results of MVD for HFS inother series.
Barker et a14 reported long term results forMVD with 86% complete and 5% partialresponse rates (> 75% relief of symptoms),with a mean follow up of eight (range 1-20)years. Iwakuma et al reported a series of 74patients with a follow up ranging from onemonth to three years.26 Seventy two (97%) hadcomplete relief, one patient partial relief, inone patient there was no benefit, and therewas one recurrence.
Panagopoulos et aTh performed the operationon 29 patients with HFS. The follow up in thisseries ranged from six months to 8-5 years, nomean value being mentioned. Twenty sixpatients (90%) had complete relief of theirspasms and three failed to benefit.
Piatt and Wilkins37 described a series of 48patients in whom 62-5% had an excellent and25% a good response. There were six failures orrecurrences. Forty one patients followed up for amean of 8-1 years showed 30 (73%) patientswith a continuing excellent response and six(16%) with a good response. Only five patientshad significant residual or recurrent HFS.38
Auger et aP reported a series of 54 cases,followed up for a mean of 3-9 years, with 44patients having a complete and five a partialresponse. Five patients had no benefit and insix patients the spasms recurred.By comparison, we report the data on 83
cases, of whom 78 were available for long-term follow up with a mean of 8 0 years.Complete relief of symptoms was obtained in72 (92-5%) and two had partial relief. In onecase there was no benefit and three patientshad a recurrence (table 5).
75
on March 9, 2020 by guest. P
rotected by copyright.http://jnnp.bm
j.com/
J Neurol N
eurosurg Psychiatry: first published as 10.1136/jnnp.60.1.72 on 1 January 1996. D
Table 6 Complications encountered in published series for patients undergoing MVD for HFS
No of No ofpatients Total TotalNo i'l comtipleted with nio penrnanent temilporarsv Seventh nerve Seventh nerv)1e Eighth 1zei zue EiglthinerveAnithlor series operationis comiplicationls conmplicaticonzs comiiplicationzs permanent temllporamy penrmanent tenporary
Auger etali 54 NA 35 (65(X)) 15 (28%,,)* 18 (33%)* 2 (4%) 9 (17%,) 8 (15) NABarker etal-I 648 648 NA NA NA 22 (34%Y.) 25 (39%) 21 (3 3%) 0Iwakuma et al` 74 NA NA NA NA NA NA 12 (16%)Loeser and Chen 20 21 12 (57(X)) 5 (24(X)) 4 (19%o) 1 (4 8) 1 (4 8%) 4 (192%o)(personal series)Loeser and Chen" 433 450 341 (76Y1o) 71 (16(I1) 38 (81o) 26 (6%1o) v58 (13%1o)(literature review)Panagopoulos et al" 29 NA 21 (72.5%1o) 8 (260/%)* 5 (17(Yo) 1 (35%) 3 (291)) 4 (14%() 1 (3 5%()Piatt and Wilkins, 48 NA 28 (60 5%) 6 (12 5%) 14 (30%) 0 2 (4%) 2 (4%(o) 1 (2%()Illingworth et al (this paper) 83 85t 57 (67 1%) 8 (9 4%) 20 (23.511) 1 (1 2111) 9 (10 6%X>) 7 (8 3%)t 4 (4 7%/)%*A patient may have had more than one post-operative complication.tOne procedure abandoned.t3°/% in the last 63 patients.l1 6'S} in the last 63 patients.NA = not available.
MVD for trigeminal neuralgia was followedby a long term recurrence rate of 5% per year inthe series of Burchiel et al39 and this finding issupported by Breeze and Ignelzi,'" who noteda 13% recurrence in a series of 52 patients fol-lowed up for an average of 23 (range 1-53)months. In our series of patients with HFStreated by MVD a late recurrence rate has notbeen found.
Table 6 describes the complicationsencountered in series published to date. Themost frequent complication was damage to theseventh and eighth cranial nerves within thecerebellopontine angle.
Hanakita and Kondol" reported the seriouscomplications encountered in a series of 278patients who underwent MVD, of which 239operations were for HFS. They encounterednine complications including an acute intrac-erebellar haematoma, delayed cerebellarswelling, status epilepticus due to supratentor-ial air, and immediate brainstem infarction,none of which were found in our series. Usingintraoperative monitoring of auditory brainstem evoked potentials, the rate of postopera-tive hearing disturbance was 4% in this series.Other complications were not discussed.
In our series there was a temporary compli-cation rate of 23-5%. A permanent neurologicaldeficit occurred in eight (9 4%) patients, thecommonest being unilateral sensorineuraldeafness.
Three quarters of the complicationsoccurred in the first 20 patients. In the other63 patients the total, comprising both perma-nent and temporary complications, fell to11%, by contrast with 32-9% for the entireseries. As experience increased it was realisedthat the only significant vascular compressionoccurred at the root exit zone. This meant thatthere was no need to explore the full subarach-noid course of the seventh nerve, and muchless dissection was required. The operationtherefore can be very precisely targeted toexpose the root exit zone of the seventh nerveas described earlier. The importance of thisapproach is that it brings the surgeon directlydown on to the root exit zone and because theseventh and eighth nerves are approachedtransversely to the line of the nerves. Thisapproach, using a narrow retractor blade to liftthe cerebellum and choroid plexus off theninth and tenth nerves, therefore avoids any
axial traction on the seventh and eighthnerves. It is this axial traction which can cause
deafness, and the approach described shouldavoid this distressing complication. This find-ing is supported by the fact that in our seriesthe permanent deafness rate fell to 3% and thepartial deafness rate to 1 6% in the final 63patients.
Further refinements suggested by Jannettaare the intraoperative monitoring of audio-graphic responses to protect against inadver-tent eighth nerve damage,-2 and by monitoringthe facial EMG, and continuing the operationuntil lateral spread responses are eliminated toimprove the cure rate.43 Neither of us hadbrain stem auditory evoked potentials or facialEMG available during the series. We thinkthat evoked potentials are only of limited valueto protect hearing because of the time neces-
sary to summate multiple responses, and thatthe method is no substitute for the surgicaltechnique we have described. The surgicalprocedure requires complete decompressionof the seventh nerve root entry zone by relievingvascular compression. The use of intraopera-tive facial nerve EMG does change thisrequirement. However desirable these moni-toring techniques may be we do not think thatthey are mandatory. The high level of relief ofHFS and the low incidence of deafness in thelast 63 patients once the technique was refinedare compatible with other series (tables 5 and6) and, we think, support our opinion.The tensor tympani and stapedius muscles
are innervated by the facial nerve. When anacoustic stimulus is applied to the contralat-eral ear, the stapedius muscle contracts andcompliance decreases. This is the acoustic-stapedial reflex. Diamant et al-"4 reported twopatients in whom stapedius muscle contrac-tions occurred synchronously with the facialtwitches. Kim and Fukushima reported theresults of impedance audiometery in a groupof 15 patients who had simultaneous "tym-panic noise"-which may be the same as thedrumming described by four of our patients-and HFS.') The abnormalities foundnamely, tonic contraction of the stapediusmuscle during tonic facial spasm and anabsence of the stapedial reflex were abol-ished by surgery. In a similar manner to theseries of Kim and Fukushima, '5 all ourpatients were free of this distressing symptom
on March 9, 2020 by guest. P
rotected by copyright.http://jnnp.bm
j.com/
J Neurol N
eurosurg Psychiatry: first published as 10.1136/jnnp.60.1.72 on 1 January 1996. D
Heinifacial spasm: a prospective long term follow up of 83 cases treated by microvascular decomnpression at two neurosurgical centres in the UK
after surgery and have remained so at the timeof follow up.MVD of the facial nerve performed by expe-
rienced surgeons using the correct approach tothe root exit zone is a relatively safe procedureand gives the patient the only chance of perma-nent relief. It should be offered to all patientswith HFS, who are deemed fit for surgery. Thealternative therapeutic option of botulinumtoxin should be available for patients who areunfit or unwilling to undergo major surgery.
These two methods are very differentapproaches to the treatment of HFS. Bothhave advantages and disadvantages. Ideally allpatients with this condition should be providedwith sufficiently detailed information to allowthem to make a well informed and considereddecision.
1 Auger RG, Whisnant JP. Hemifacial spasm in Rochesterand Olmsted county, Minnesota, 1960 to 1984. ArchNeuirol 1990,47: 1233-4.
2 Jannetta PJ, Abassy M, Maroon JC, Ramas FM, Albin MS.Etiology and definitive microsurgical treatment of hemi-facial spasm. Operative techniques and results in 47patients. _7 Nezurosuirg 1977;47:321-8.
3 Auger RG, Piepgras DG, Laws ER. Hemifacial spasmresults of microvascular decompression of the facial nervein 54 patients. Mayo Clin Proc 1986;61:640-4.
5 Maroon JC, Lunsford LD, Deeb ZL. Hemifacial spasmdue to aneurysmal compression of the facial nerve. ArchNeurol 1978;35:545-6.
6 Miyazaki S, Fukushima T. CP angle epidermoid presentingas hemifacial spasm. No To Shinkei 1983,35:951-5.
7 Pierry A, Cameron M. Clonic hemifacial spasm from poste-rior fossa arteriovenous malformation. _7 Neurol NeurosuirgPsychiatry 1979;42:670-2.
8 Jannetta PJ. Trigeminal neuralgia and hemifacial spasm:etiology and definitive treatment [abstract]. Arch Neuirol1975;32:353.
9 Jannetta PJ. Hemifacial spasm. In: Samii M, Jannetta PJ,eds. The cranial nerves: anatomy, pathology, pathophysiol-ogy, diagnosis, treatmient. New York: Springer-Verlag,1981:484-93.
10 Loeser JD, Chen J. Hemifacial spasm: treatment by micro-surgical facial nerve decompression. Neurosurgery 1983;13: 141-6.
11 Jannetta PJ. Treatment of trigeminal by suboccipital andtranstentorial cranial operations. Cliuz Neuirosurg 1976;23:538-49.
12 Sugita K. Microneurosuirgical atlas. Berlin: Springer-Verlag,1985:237.
13 Perkin GD, Illingworth RD. The association of hemifacialspasm and facial pain. _7 Neurol Neurosuirg Psychiatry1989;52:663-5.
14 Alexander GE, Moses H. Carbamazepine for hemifacialspasm. Neuirology 1982;32:286-7.
15 Herzberg L. Management of hemifacial spasm with clon-azepam. Neurology 1985;35:1676-7.
16 Hughes EC, Brackmann DE, Weinstein RC. Seventh nervespasm: effect of modification of cholinergic balance.Otolaryngol Head Neck Surg 1980;88:491-9.
17 Mauriello JA, Aljian J. Natural history of treatment of facialdyskinesias with botulinum toxin: a study of 50 consecu-tive patients over seven years. Br_ Ophthlmnol 1991;75:737-9.
18 Yoshimura DM, Aminoff M, Tami TA, Scott AB.Treatment of hemifacial spasm with botulinum toxin.Muiscle Nerve 1992;15:1045-9.
19 Elston JS. The management of blepharospasm and hemifa-cial spasm. _7 Neuirol 1992;239: 5-8.
20 Kalra HK, Magoon EH. Side effects of the use of botu-linum toxin for treatment of benign essential ble-pharospasm and hemifacial spasm. Ophlthtalmlic Sinrg1990;21:335-8.
21 Taylor JDN, Kraft SP, Kazdan MS, et al. Treatment of ble-pharospasm and hemifacial spasm with botulinum Atoxin: a Canadian multicentre study. Can _7 Ophthalmol1991;26:133-8.
22 Apfelbaum RI. A comparison of percutaneous radiofre-quency trigeminal neurolysis and microvascular decom-pression of the trigeminal nerve for the treatment of ticdouloureux. Neurosurgery 1977; 1:16-22.
23 Harris W, Wright AD. Treatment of clonic facial spasm:(a) by alcohol injection. (b) by nerve anastomosis. Lanicet1932;i:657-62.
24 Andrew J. Surgery for involuntary movements. BrY HospMed 1981;26:522-8.
25 Wakasugi B. Facial nerve block in the treatment of facialnerve spasm. Arch Otolaryngol 1972;95:356-8.
26 Iwakuma T, Matsumoto A, Nakamura N. Hemifacialspasm: comparison of three operative procedures in 110patients. _7 Neutrosurg 1982;57:753-6.
27 Fisch U, Esslen E. The surgical treatment of facial hyperki-nesia. Arch Otolaryn,zol 1972;95:400-4.
30 Moller AR, Janetta PJ. On the origin of synkinesis in hemi-facial spasm: results of intracranial recordings. _7Neurosurg 1984;61 :569-76.
31 Nielsen VK, Jannetta PJ. Pathophysiology of hemifacialspasm: effects of facial nerve decompression. NVeurolo,gy1984;34:891-7.
32 Moller AR. Interaction between the blink reflex and abnor-mal muscle response in patients with hemifacial spasm:results of intraoperative recordings 7 Neurol Sci 1991;101:114-23.
33 Haines SJ, Torres F. Intraoperative monitoring of the facialnerve during decompressive surgery for hemifacialspasm. . Neurosurg 1991;74:254-7.
41 Hanakita J, Kondo A. Serious complications of micro-vascular decompression operations for trigeminalneuralgia and hemifacial spasm. NeurosurgeD, 1988;22:348-52.
42 Moller AR, Jannetta PJ. Monitoring auditory functionsduring cranial nerve microvascular decompression opera-tions by direct recording from the eighth nerve. 7Neurosuirg 1983;59:493-9.
44 Diamant H, Enfors B, Wilberg A. Facial spasm. With specialreference to the chorda tympani function and operativefunction. Laryngoscope 1967;77:350-8.
45 Kim P, Fukushima T. Observations on synkinesis inpatients with hemifacial spasm. Effects of microvasculardecompression and aetiological considerations. .Neuirosuirg 1984;60:821-7.
77
on March 9, 2020 by guest. P
rotected by copyright.http://jnnp.bm
j.com/
J Neurol N
eurosurg Psychiatry: first published as 10.1136/jnnp.60.1.72 on 1 January 1996. D
