Main Article

Facial palsy secondary to cholesteatoma: analysis ofoutcome following surgery

M A SIDDIQ, L M HANU-CERNAT, R M IRVING

AbstractFacial palsy is a rare presenting feature of cholesteatoma. Prompt treatment usually results in a goodoutcome, but if treatment is delayed the prognosis can be difficult to predict. We retrospectivelyreviewed 326 consecutive patients who had undergone temporal bone surgery for cholesteatoma.Eleven patients had presented with facial palsy, of whom eight had petrous apex involvement and threehad disease confined to the middle-ear cleft. All patients with middle-ear disease were operated uponwithin two months of presentation, and all showed some recovery in facial function. The diagnosis hadbeen delayed in all eight cases of apical disease, with four cases having had a long-standing totalweakness. One case treated after seven months’ partial weakness achieved a full recovery. In threecases of long-standing partial weakness, pre-operative facial function was preserved by maintaining thefacial nerve in its normal anatomical location.

Facial palsy associated with cholesteatoma should be treated as early as possible. Recovery can occureven if treatment is delayed for up to seven months. After this time, recovery is increasingly unlikely,but any residual function can be maintained and further deterioration prevented.

Key words: Cholesteatoma; Facial Paralysis; Ear, Middle; Petrous Bone

Introduction

Facial nerve paralysis is a rare but recognized compli-cation of cholesteatoma, of which it can unusually bea presenting feature. Its prevalence has decreasedfrom 1–2.3 per cent of cholesteatoma presentationsin the 1950s1,2 to 0.04–0.16 per cent of thosetoday.1,3,4 This reduction is most probably due toearlier diagnosis and treatment.

Despite numerous reports in the literature, there isstill little information on rates of recovery from facialnerve palsy in relation to duration and severity ofparalysis. The objective of the present study wastherefore to develop prognostic information for therecovery of facial nerve paralysis associated withcholesteatoma, by examining the post-surgical out-come. This could potentially guide surgical decision-making regarding the urgency of intervention andthe operative management of the facial nerve, andit could also help to counsel similarly presentingpatients in the future.

Materials and methods

A five-year, retrospective analysis of the medicalrecords of all patients presenting with cholesteatomato our tertiary referral centre (the department

of otorhinolaryngology, head and neck surgery,University Hospital Birmingham) was carried out.We selected all patients with a diagnosis of cholestea-toma who had presented with facial palsy.

A total of 326 patients underwent temporalbone surgery for cholesteatoma during the periodJanuary 1998 to December 2003. Eleven of thesepatients (3.4 per cent) presented with facial nervepalsy. Of the 326 patients, 16 had disease involvingthe petrous apex, of whom eight presented withfacial palsy (50 per cent). The remaining 310 patientshad disease confined to the middle-ear cleft, of whichthree presented with facial palsy (0.93 per cent).

All patients presenting with facial paralysis under-went surgery conducted by the senior author (RI). Inthe cases with petrous apex disease, all eight patientshad a dead ear, and the surgical approach wastrans-otic, with preservation of the facial nerve inits fallopian canal in those with some residual facialfunction or a recent onset of paralysis. In cases oflong-standing paralysis, a total petrosectomy wascarried out without facial nerve repair.

A radical or modified radical mastoidectomyapproach, using a small cavity ‘front to back’ tech-nique, was used in the three cases with middle-ear disease. The nerve in these three cases was

From the Department of Otorhinolaryngology, University Hospital Birmingham, Birmingham, UK.Accepted for publication: 27 June 2006.

The Journal of Laryngology & Otology (2007), 121, 114–117.# 2006 JLO (1984) Limiteddoi:10.1017/S0022215106003227Printed in the United KingdomFirst published online 2 November 2006

114

maintained in its normal anatomic location andwidely decompressed at the site of involvement bycholesteatoma.

We recorded any previous surgery, the duration ofpre-operative facial nerve palsy, and the grade of thefacial nerve paralysis both pre- and post-operatively(according to the House–Brackmann gradingsystem).5 In addition, the facial nerve segmentaffected by disease and the site of the cholesteatoma(whether apical or limited to the middle-ear cleft)were recorded.

Results

Of those patients presenting with facial palsy, sevenwere male and four female, with a mean age of56 years (range 30–80 years).

Among the presenting features, all patients hadhearing loss, one patient had suffered four previousepisodes of meningitis confirmed by lumbar punctureand two patients had evidence of an abducens nervepalsy (one of whom had experienced three previousepisodes of abducens nerve palsy which had allrecovered spontaneously).

The duration of pre-operative facial palsy rangedfrom one month to 21 years. In eight cases, theparalysis was of sudden onset. Two cases weregradual losses. The first had undergone a mastoidexploration 14 years previously and presented withan 18 month history of progressive palsy which wasgrade VI on presentation. The second patient alsohad a mastoid exploration for cholesteatoma threeand a half years ago but post-operatively progressedfrom grade IV to grade V facial paralysis, at whichtime he was referred to our department. Onepatient had a relapsing loss. He initially had asudden facial paralysis which partly recovered butthen progressed to a grade VI palsy. He had under-gone a modified radical mastoidectomy followingthis but presented to us with residual disease and agrade VI palsy. All three of these patients haddisease involving the petrous apex.

Patients with apical bone involvement had longerperiods of established pre-operative facial palsy(Table I).

The tympanic segment of the facial nerve wasthe commonest site of nerve involvement in themiddle-ear cleft cases. In cases with apical involve-ment, the labyrinthine and geniculate segmentswere most commonly involved, although more thanone site was affected in six cases.

Of the patients with apical disease, we observedthat only one showed some improvement in facialnerve function post-operatively (Figure 1), whilethe remaining seven patients showed no deterio-ration or improvement. However, all three patientswith middle-ear disease showed some post-operativeimprovement in facial nerve function. Additionally,the patients with middle-ear disease had all pre-sented within two months of their facial palsy,whereas this time period had been up to 21 years inpatients with apical disease.

The post-operative progress of facial nerve func-tion was monitored during a follow-up period thatranged from one month to five years.

Discussion

In those patients with petrous apex cholesteatoma,50 per cent (eight of 16) presented with facial nervepalsy. In the literature, the incidence of facial palsyin petrous apex cholesteatoma is reported to bebetween 50 to 81 per cent.6 – 9 The duration of facialpalsy in these cases tends to be several years, asreported by several authors.6,7,10 Additionally, thechance of recovery of facial nerve function post-operatively is reported to be small, and, althoughthe majority of patients maintain facial nerve func-tion post-operatively, there are reports of deterio-ration in function.6,7 In our series of eight caseswith apical involvement, the duration of facial palsyhad been over 3.5 years (range 3.5 to 20 years)in seven cases, and none of these had shown anypost-operative improvement in function (House–Brackmann grading had remained at grade III in

TABLE I

DETAILS OF FACIAL PALSY

Patient no Previous surgery Facial palsyduration

Facial palsy grade Cholesteatoma site Facial n segmentinvolved

Pre-op Post-op

1 None 20 yrs III III Apical L, G2 Mastoid exploration 21 yrs III III Apical L, G, T, M3 Mastoid exploration 4 yrs VI VI Apical T, G4 None 10 yrs VI VI Apical L, G5 None 7 mths II I Apical L, G, M6 Modified radical

mastoidectomy7 yrs VI VI Apical L

7 None 2 mths IV II Middle ear T8 None 1 mth III II Middle ear T9 Modified radical

mastoidectomy3.5 yrs V V Apical T

10 None 20 yrs VI VI Apical L, G11 None 1 wk IV II Middle ear T

No ¼ number; Pre-op ¼ pre-operative; Post-op ¼ post-operative; n ¼ nerve; yr ¼ year; mth ¼ month; wk ¼ week;L ¼ labyrinthine; G ¼ geniculate; T ¼ tympanic; M ¼ mastoid

FACIAL PALSY SECONDARY TO CHOLESTEATOMA 115

two cases, at grade V in one case and at grade VI inone case) (cases one to four, six, nine and 10 inTable I). However, importantly, none of the eightcases had shown any deterioration in function. Inone case (case five in Table I), however, the durationof facial palsy was seven months and an improvementfrom grade II to grade I was seen.

We define apical disease as that involving bonemedial to the bony labyrinth. In these cases, thedisease could have originated as a congenital choles-teatoma within the apex and spread laterally toinvolve the facial nerve. Alternatively, the observeddisease could represent medial spread of a congenitalor acquired lesion arising within the middle-ear cleft.In apical disease, it is not always possible to deter-mine the exact site of origin of the cholesteatoma.Also, it may be difficult to determine whether ornot cholesteatomas are congenital in origin, althoughthe majority in this series most probably representtrue congenital, apical cholesteatomas.

The surgical approach to the facial nerve in thesecases varies. The approaches include anterior andposterior mobilization,6 decompression and rerout-ing of the nerve,7 sacrificing the nerve with an

interposition sural nerve graft or facial–hypoglossalanastamosis,8 and preserving the nerve within thefallopian canal (our preferred method). In ourseries, there were no cases of an iatrogenic post-operative facial palsy, and, indeed, none of the 16patients with petrous apex cholesteatoma haddeterioration of facial nerve function post-operatively. However, anterior and posterior mobil-ization,6 rerouting7 and interposition grafting8 haveall resulted in some cases of deterioration of facialnerve function.

We found the commonest site of facial nerveinvolvement in apical disease to be the labyrinthinesegment and the geniculate ganglion. The literaturesupports the geniculate ganglion6,7 area as beingthe most commonly affected site in apical disease.

The incidence of facial nerve palsy presentingwith middle-ear cleft cholesteatoma in our seriesis 0.93 per cent (three of 310). Again, there wereno cases of acquired post-operative facial palsy inthe 310 cases, and indeed all three cases showedimprovement in facial nerve function post-operatively. In the literature, the incidence of facialpalsy in cholesteatoma confined to the middle-earcleft is 0.002–2.6 per cent.11 – 13Although grading isnot widely reported, the incidence of completerecovery of facial nerve function is 60–84 percent.12 – 14 Most reports do not mention the timingof surgery. However, Yetiser et al.11 reported11 patients with facial nerve palsy and middle-earcholesteatoma, all of whom were operated uponwithin 30 days, with complete recovery of facialnerve function in seven cases and partial recoveryin four cases. In addition, Ikeda et al.15 reviewed 16cases and found that the outcome had been betterin those operated upon within two months (13 of 16cases). In our series, the three middle-ear cleftcases were operated upon within two months; theirpre-operative House–Brackmann status weregrades IV, III and IV (cases seven, eight and 11 inTable I, respectively), and they all improved tograde II post-operatively.

. Facial nerve palsy is a rare presenting featureof cholesteatoma, occurring in 50 per cent ofapical and 0.93 per cent of middle-ear cases inthis series

. In cases of cholesteatoma complicated bypartial facial weakness, prompt surgery isrecommended. Improvement in facial nervefunction can however be achieved even ifsurgery is delayed for up to seven months

. A partial facial weakness present for morethan three and a half years is unlikely toimprove post-operatively and is associatedwith petrous apex disease

. Petrous apex cholesteatomas can be surgicallyremoved with no detriment to facial nervefunction by maintaining the nerve in its normalanatomical location

FIG. 1

CT images from a case of limited apical extension ofcholesteatoma, demonstrating (a) involvement of the labyrin-thine segment of the facial nerve (large labyrinthine fallopiancanal) and (b) eroded/absent anterior limb of the superiorsemicircular canal. This patient presented with a dead earand a grade II facial palsy which improved to grade I following

surgery.

M A SIDDIQ, L M HANU-CERNAT, R M IRVING116

For middle-ear cleft cholesteatoma again we findthe surgical approach to the facial nerve varies.Limited decompression of the oedematous areaof the fallopian canal,15 complete decompressionof the canal from geniculate ganglion to stylo-mastoid foramen without nerve sheath incision,11

and retention of the nerve in its canal withoutdecompression have been described.14 We performdecompression of the involved segment of thenerve only.

The commonest reported affected segment of thefacial nerve in middle-ear cleft disease is the tympa-nic segment.1,11,12 This was also found to be thecase in all three of our patients with middle-earcholesteatoma.

Conclusions

In cases of cholesteatoma complicated by facialweakness, prompt surgical treatment is recom-mended. Post-operative improvement in facialnerve function can be anticipated even if the paraly-sis has been present for up to seven months.Recovery may occur in more long-standing cases.However, if the facial palsy has been present formore than three and a half years, our data wouldsuggest that post-operative improvement in facialnerve function would be exceptional. In apicaldisease complicated by long-standing partial weak-ness, pre-operative facial function can be preservedand further deterioration prevented by maintainingthe facial nerve in its normal anatomical location.

References

1 Savic DJL, Djeric DR. Facial paralysis in chronic suppura-tive otitis media. Clin Otolaryngol 1989;14:515–17

2 Harker LA, Pignatari SSN. Facial nerve paralysis second-ary to chronic otitis media without cholesteatoma. Am JOtol 1992;13:372–4

3 Pollock RA, Brown LA. Facial paralysis in otitis media. In:Graham MD, House WF, eds. Disorders of the FacialNerve. New York: Raven Press, 1982;221–4

4 Palva T, Virtanen H, Makinen J. Acute and latent mastoi-ditis in children. J Laryngol Otol 1985;99:127–36

5 House JW, Brackmann DE. Facial nerve grading system.Otolaryngol Head Neck Surg 1985;93:146–7

6 Axon PR, Fergie N, Saeed SR, Temple RH, Ramsden RT.Petrosal cholesteatoma: management considerations forminimizing morbidity. Am J Otol 1999;20:505–10

7 Atlas MD, Moffat DA, Hardy DG. Petrous apex cholestea-toma: diagnostic and treatment dilemmas. Laryngoscope1992;102:1363–8

8 Magliulo G, Terranova G, Sepe C, Cordeschi S, Cristofar P.Petrous bone cholesteatoma and facial paralysis. ClinOtolaryngol 1998;23:252–8

9 Cannoni M, Pech A, Fuchs S, Zanaret M, Thomassin JM,Dessi P et al. Congenital cholesteatoma of the petrousbone. Etiopathogenic discussions apropos of 11 cases.Rev Laryngol Otol Rhinol 1989;110:33–42

10 Yanagihara N, Matsumoto Y. Cholesteatoma of thepetrous apex. Laryngoscope 1981;91:272–8

11 Yetiser S, Tosan F, Kazkayasi M. Facial nerve paralysis dueto chronic otitis media. Otol Neurotol 2002;23:580–88

12 Altuntas A, Unal A, Aslan A, Ozcan M, Kurkuoglu S,Nalca Y. Facial nerve paralysis in chronic suppurativeotitis media: Ankara Numune hospital experience. AurisNasus Larynx 1998;25:168–72

13 Osma U, Cureoglu S, Hosoglu S. The complications ofotitis media: report of 93 cases. J Laryngol Otol 2000;114:97–100

14 Baljosevic I, Micic S, Baljosevic Z, Milovanovic J. Facialnerve paralysis as a sequelae of chronic suppurative otitis[in Croatian]. Med Pregl 2000;53:93–6

15 Ikeda M, Nakazato H, Onoda K, Hirai R, Kida A. Facialnerve paralysis caused by middle ear cholesteatoma andeffects of surgical intervention. Acta Otolaryngol 2006;126:95–100

Address for correspondence:Mr M A Siddiq,72, Launceston Road,Walsall,West Midland WS5 3EE,UK.

Fax: 01922 613172E-mail: azhersiddiq@hotmail.com

Mr M A Siddiq takes responsibility for the integrity of thecontent of the paper.Competing interests: None declared

FACIAL PALSY SECONDARY TO CHOLESTEATOMA 117