optic nerve glioma and the of optic nerve tumours...

British Journal of Ophthalmology, 1989, 73,967-974

Optic nerve glioma and the managementof optic nerve tumours in the youngJ E WRIGHT, A A McNAB, AND W I McDONALD

From the Orbital Clinic, Moorfields Eye Hospital, City Road, London EC]V2PD

SUMMARY Thirty-one patients presenting as orbital optic nerve glioma have been reviewed withmaximum follow-up of 14 years. Sixteen of these patients have been reported on previously andfurther follow-up is provided. Sixteen patients had a stable clinical course with little change over aperiod of up to 13*5 years. Neurofibromatosis was relatively common in this group (11/16). Fifteenpatients had progressive enlargement of the tumour; the incidence of neurofibromatosis in thisgroup was low (4/15). Eleven of these patients were explored neurosurgically and the optic nervetotally excised in 10 of them. The proximal cut end was normal in six patients and the chiasm hasapparently remained free oftumour in all of them. We suggest a method of management of primaryoptic nerve tumours, both meningiomas and gliomas, in young patients.

There is continuing debate about the natural historyand management of optic nerve gliomas. Hoyt andBaghdassarian' and others2' have claimed that thesetumours behave as hamartomas and that they shouldbe managed conservatively. Others have reacheddifferent conclusions, stating that these tumours arenot benign and self-limiting, and that treatment isrequired." In 1980 we reported on 17 patients anddemonstrated a division into one group which wasstable and another which showed active growth.5 Oneof these patients was subsequently found to have ameningioma, not a glioma as was first reported, and isdescribed as case 1 in the accompanying paper.8 Sincethen we have seen a further 15 patients and they toohave fallen into one of these two groups. We haveadopted a radical surgical approach in those patientswith evidence of active growth and without evidenceof chiasmal involvement. In this paper we report ourresults and follow-up of 31 patients with this un-common condition.

Patients and methods

Since 1973 we have seen 31 patients with an initialdiagnosis of optic nerve glioma. The diagnosis wasmade on the basis of typical clinical and radiologicalfindings, and in 16 patients in whom diagnostic tissue

Correspondence to J E Wnght, FRCS, Moorfields Eye Hospital,City Road, London EC1V 2PD.

was obtained our clinical diagnosis was confirmedhistopathologically.We have excluded the small number of older adults

with the highly malignant form of optic glioma(glioblastoma), which usually involves the intra-cranial visual pathways and follows a quite differentcourse.9A detailed clinical and family history was obtained

and a full ocular and neurological examinationperformed in each patient. Visual fields werecharted whenever possible, and most patients hadpattern visual evoked potentials (VEP) recorded.Patients were carefully examined for stigmata of vonRecklinghausen's neurofibromatosis by the criteriaof Crowe and Schull."' Plain skull films with opticcanal views were routinely performed, and since 1975all patients have had CT scans at presentation. Inrecent years most patients have also undergonemagnetic resonance imaging (MRI).The patients were assigned to two groups, (1) those

with a 'stable' course and (2) those with 'activegrowth', on the basis of changes in vision and/orproptosis. A change was indicated by either thehistory (patient's or relatives' observations) orfollow-up examination. For inclusion in the stablegroup our criteria were: (a) unchanged or improvedvision, or vision worsening by less than 2 lines over aminimum period of two years, and (b) proptosisunchanged, decreasing or increasing by 2 mm or lessover a minimum of two years. For inclusion in the

967

on 7 July 2018 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.73.12.967 on 1 Decem

ber 1989. Dow

nloaded from

http://bjo.bmj.com/

J E Wright, A A McNab, and W I McDonald

active growth group the criteria were (a) visual loss atthe rate of more than 2 lines in a year, or (b) a rate ofincrease in the proptosis of more than 2 mm in a year.This leaves a potential gap for 'intermediate' patients.None of our patients qualified for this category.While the patients had a spectrum of change in

vision and/or acuity, they fell clearly into one or othergroup. For example, patients with active growthusually had a dramatic increase of proptosis overperiods of less than a year, whereas those in the stablegroup remained unchanged or even improved over

many years.

Follow-up was every six to 12 months in the groupwith an established stable course but every two to sixmonths in the group with evidence of active growth.

Clinical, radiological, and electrophysiologicalinvestigations were aimed at determining theanatomical extent of the tumour and in particularwhether the optic chiasm was involved. Thosepatients with active growth, poor vision, and noevidence of chiasmal disease were referred to aneurosurgeon for a frontal craniotomy. If the chiasmappeared macroscopically normal, the nerve was

transacted just in front of the chiasm, and afterunroofing the optic canal and orbital roof the involvednerve was removed to the globe.

In patients in whom there was active growth butinvestigations revealed an involved chiasm surgery,when required, was confined to the reduction ofthe disfiguring proptosis by means of an anteriorapproach, or a ventriculo-atrial shunt for obstructivehydrocephalus.

Biopsy of the optic nerve has not been performedfor many years because the results can be misleading.'Radiotherapy was not used routinely, and only twopatients have had this form of treatment. Squintsurgery was occasionally performed for cosmeticpurposes and enucleation was required in severalcases.

Results

CLINICAL FINDINGSA total of 31 patients has been seen with a clinicaldiagnosis of optic nerve glioma. The details of theirage and sex distribution are shown in Table 1. Thepresenting symptoms are outlined in Table 2 and thephysical findings in Table 3. Fifteen of the 31 patientshad neurofibromatosis (48%). All had reduced acuityat presentation, with 15 having 6/60 or worse vision(Table 4). the average degree of proptosis was 3-0mm, but six had no measurable proptosis whereasone was proptosed 8 mm. All but one patient had a

relative afferent pupil defect, and this patient was

found to have bilateral optic nerve involvement on

CT but had presented with unilateral proptosis and

visual failure. The ocular motility disturbance mostcommonly seen (14/31) was a limitation of elevationof the globe, which in many cases was quite marked.

RADIOLOGYAn enlarged optic foramen on plain x-ray was foundin 25 patients at presentation. Twenty-six patientshad CT scans at presentation (Fig. 1); 24 showed amass involving the optic nerve without apparentinvolvement of the optic chiasm. Two showedobvious chiasmal enlargement and two incidentalenlargement of the contralateral optic nerve. Thesechanges were unsuspected on the basis of clinicalexamination.

Magnetic resonance imaging (MRI) has been

Table 1 Age and sex data

Total patients 31Sex Male 7 Female 24Involved nerve Right 15 Left 16Age at presentation

Mean 10(2 yearsMedian 6-5 yearsRange 2-38 years

Table 2 Presenting symptoms

Proptosis only 12Proptosis and visual loss 9Visual loss only 3Proptosis and squint 3Squint and visual lossSquint onlyPain, visual loss, and obscurations on

lateral gaze

Table 3 Physicalfindings

Present Absent

Poor vision 31Proptosis 25 6Relative afferent pupil defect 30 1Enlarged optic canal 24 7Neurofibromatosis 15 16

Table 4 Visual acuity

Vision Nutmber

6/9-6/12 56/18-6/24 66/36-6/6( 9CF-HM 4PL 3NPL 4

CF=counting fingers. HM=hand movements. PL=perception oflight. NPL=no PL.

968


http://bjo.bmj.com

/B

r J Ophthalm


ber 1989. Dow

nloaded from

http://bjo.bmj.com/

Optic nerve glioma and the management ofoptic nerve tumours in the young

Fig. 3 Coronal MRI of the same patient in Figs. I and 2showing involvement of the optic chiasm (arrowed).

Fig. I Axialx t , witnoutcontrast, oja patent witn an activeglioma involving the right optic nerve. Note the typicalfusiform outline ofthe tumour and the enlarged optic canal(arrowed).

particularly useful for demonstrating the extent oftumour (Figs. 2 and 3). In two patients it has detectedchiasmal involvement when both CT and VEPs were

normal.

Fig. 2 Axial MRI ofthe same patient as in Fig. 1. Note theimproved definition ofthe enlarged optic nerve in the opticcanal (arrowed).

CLINICAL COURSEStable groupEight of nine patients reported in 19801 as having astable clinical course have not shown evidence ofactive growth subsequently. One patient has sincebeen found to have a meningioma and has beenexcluded from this study (patient 1 in the accom-panying paper on optic nerve meningioma'). Eightnew patients have been added to this group. Thestable group had ages at presentation ranging from2-2 to 38 years (mean 9 9 years, median 8 years), andat the onset of their symptoms from birth to 30 years(mean 5 9, median 3-5 years). Eleven of 16 hadneurofibromatosis. In three patients in whom clinicalexamination suggested the glioma was confined toone optic nerve CT scans showed thickening of thecontralateral nerve in two and the chiasm in theother. The optic disc was atrophic in 14/16 patients,with shunt vessels in one. Only two showed optic discswelling, which was mild. Visual acuity improvedspontaneously in three patients, one from 6/24 to 6/6over seven years, and minor fluctuations in visionwere common. Two patients also had a reduction intheir proptosis of 2 and 3 mm in the period of follow-up. The greatest increase in proptosis was from 3 to 6mm over six years. The remainder were unchangedor had increases of I or 2 mm over periods of two to12 years. Follow-up has ranged from two to 13-5years (mean 7-2, median 7-0 years). The exceptionwas a patient with a one-month follow-up whopresented at the age of 38 with a 27-year history ofpoor but stable vision and neurofibromatosis. Fourpatients within this group have been lost to follow-upafter 2, 2-5, 3, and 8 years. Another patient within

969


http://bjo.bmj.com

/B

r J Ophthalm


ber 1989. Dow

nloaded from

http://bjo.bmj.com/

J E Wright, A A McNab, and WI McDonald

this group with neurofibromatosis and a five-yearfollow-up has a mother (not included in this series)with a chiasmal glioma that was irradiated and hasremained essentially unchanged over at least 10years.

Active groupBy contrast there were 15 patients with evidence ofprogressive enlargement of the tumour, eight addedsince 1980.1 These patients presented between theages of 3-0 and 26-4 years (mean 10-7, median 9-0years), and their symptoms began between birth and26-0 years (mean 8-3, median 6-9 years). Only four ofthis group had neurofibromatosis. Nine had discswelling, four optic atrophy, one had tumour onthe disc, and another a total exudative retinaldetachment.Two patients from the 'active growth' group were

confirmed as having chiasmal involvement by CT.One presented in 1974 ('patient 3's) with a four-months history of squint and two months' proptosis.During follow-up she developed visual evokedpotential (VEP) evidence, and later visual fieldchanges, indicative of chiasmal involvement. A CT in1974 had failed to show chiasmal involvement but a

later higher resolution scan detected it. The secondpatient presented first in 1973, ('patient 7'5) andduring follow-up developed VEP and then perimetricevidence of chiasmal involvement. Only at this stagedid a CT scan disclose an enlarged chiasm, and apneumoencephalogram confirmed a suprasellarmass. A third patient's chiasmal disease was detectedonly on MRI.

Follow-up in the 'active growth' group has beenfrom 1-5 to 14 years (mean 7-8, median 6-2 years). Nopatient has been lost to follow-up. One died after apneumoencephalogram.

SURGERYTen patients with active growth without evidence ofchiasmal disease were referred for neurosurgicalexploration. Nine had a macroscopically normalchiasm and the optic nerve was therefore excisedfrom the chiasm to the globe. The proximal end of theexcised nerve was histologically clear of tumour in sixpatients and involved by tumour in three patients.One patient had an enlarged chiasm and the tumourwas merely biopsied. A further patient had an orbitalglioma excised through a lateral orbitotomy.Whether the tumour extended into the intracanalic-ular portion of the nerve is not known.Three patients required enucleation. One had

tumour growth into the vitreous and developedintractable glaucoma. One had massive growth ofglioma from a small amount of tumour which was leftattached to the globe after the initial craniotomy. The

globe and recurrent glioma were removed together.The third patient had multiple cystic recurrences ofthe orbital part of the glioma and ultimately requiredenucleation and marsupialisation of the cyst. Severalpatients who have had the optic nerve excised havedeveloped signs of ocular ischaemia, but all haveretained the globe. The one patient explored via alateral orbitotomy ('patient 8'5) had the optic nervecontaining a large grade 1 astrocytoma excised.There has been no evidence of recurrence 12 yearsafter surgery.Two patients with active growth and chiasmal

involvement needed debulking of the orbital part ofthe glioma for disfiguring proptosis and intermittentsubluxation of the globe.

In the group with a stable course five have hadsurgery. In two a biopsy was obtained to confirm thediagnosis. We have since discontinued this practice,as small biopsies tend to give misleading information,and large biopsies compromise vision. Three haveundergone cosmetic squint surgery. One patient withneurofibromatosis had a large plexiform neuroma inassociation with a diffusely enlarged ipsilateral opticnerve. The plexiform neuroma was debulked. Thecombination of a plexiform neuroma and a presumedoptic nerve glioma is in our experience very unusual.

H ISTOPATHOLOGYIn 14 of the 15 cases with evidence of active growthpathological tissue was obtained for histologicalexamination. Two specimens were obtained at orbit-otomy to debulk tumour in patients with knownchiasmal disease. Eleven specimens were excisedoptic nerves, and one was a biopsy obtained atcraniotomy. In 11/14 a typical grade 1 pilocyticastrocyoma was found. Arachnoid hyperplasia wasprominent in seven. Three were considered to begrade 2 astrocytomas. These three all had microscopicinvolvement of the cut proximal end of the nerve butany disease within the chiasm has remained stable.Two patients with a stable clinical course werebiopsied via orbitotomy and grade I tumours werefound. There was no relationship between neuro-fibromatosis and the pattern or grade of glioma.

FURTHER CLINICAL COURSEOf the three patients in whom tumour was present inthe proximal end of their excised optic nerve, onlyone has shown evidence of advancement of theresidual intracranial tumour, with the developmentof mild disc swelling, an inferonasal field defect, anabnormal VEP and involvement of the chiasm andthe contralateral optic nerve visible with MRI (butnot CT). These changes have remained stable overfive years, and central vision remains 6/6 seven yearsafter craniotomy. This patient also developed an

970


http://bjo.bmj.com

/B

r J Ophthalm


ber 1989. Dow

nloaded from

http://bjo.bmj.com/

Optic nerve glioma and the management ofoptic nerve tumours in theyoung

orbital recurrence from tumour left attached to theglobe and required an enucleation. No further orbitalrecurrence has occurred. In the two other patientsthere has been no clinical or radiological evidence oftumour growth within the chiasm over periods of sixand nine years, but one of these patients developedmultiple cystic orbital recurrences and finally requiredenucleation.None of the six patients in whom the chiasm was

clear after excision of the optic nerve has evidence ofrecurrent chiasmal disease after being followed upfor 2, 3, 5, 9, 10, and 10 years. One developed anorbital recurrence requiring enucleation and thenpartial exenteration for a further recurrence. Theorbit is free of tumour after five years. One patientwhose glioma was removed via a lateral orbitotomyhas had no chiasmal involvement after 12 years. Wedo not know if the canalicular part of the optic nerveis free of tumour.

Discussion

The natural history and management of optic nervegliomas remains controversial. Some authors'-believe them to behave like hamartomas which areincapable of extension along the nerve into thechiasm and therefore offer no threat to the vision ofthe other eye. Worsening vision or involvement ofthe contralateral field is attributed to an increase inthe pressure of a pre-existing lesion on near-by axonsrather than evidence of true extension into previouslyuninvolved tissue. We disagree with this concept.The evidence from our present study supports theviews we expressed in 1980.' At that time we con-cluded that patients with optic nerve glioma fell intotwo fairly distinct groups, one with an indolent courseand little change in the size of the tumour, and theother with progressive tumour enlargement. Havingfollowed up these patients for a longer period andadded new patients to the series, our conclusionsremain the same. Two papers have recently addedsupport for this view. Alvord and Lofton' reviewedthe literature and applied actuarial analysis to 623cases of optic gliomas of which 155 occurred in theoptic nerve. They concluded that gliomas have a verywide but continous range of growth rates; some growrapidly probably by simple exponential doubling at aconstant rate, but the majority behave as thoughgrowth decelerates. Wong et al" have published areanalysis of the University of California, SanFrancisco, experience. Several of their patients hadbeen the subject of Hoyt's studies,' but they reacheda different conclusion, namely, that optic gliomas arenot benign, self-limiting lesions and should betreated.

Deciding whether patients had an inactive or active

lesion was usually straightforward. Most patients hadan initial period of growth of the lesion with pro-gressive proptosis. A minority of those with aninactive lesion had little or no proptosis at any stage,and poor vision was noted either accidentally or atroutine school examination. The majority progressedto a period of quiescence, which we have observed tolast a long as 13 years, with one patient reporting aninterval of 27 years. Some while under observationexperienced a reduction of proptosis and improvedvisual acuity. Active progressive gliomas continuedto enlarge, with worsening vision and increasingproptosis usually accompanied by swelling of the opticdisc andCT orMRI evidence ofgrowth of the tumour.The frequency of disc swelling and restriction of

ocular movement in patients with active lesions isprobably due to enlargement of the tumour. Ninepatients had disc oedema compared with only two inthe inactive group, where optic atrophy was muchmore common (14/16). Similarly, restriction of move-ment of the eye, usually limitation of elevation, waspresent in 11 patients with active growth, whereasonly three patients with a stable course had anyreduction of movement of the globe.The most striking difference between the two

groups was the higher incidence of neurofibromatosisin patients with a stable course (11/16) compared withthose with active growth (4/15). The tendency forgliomas of the optic pathways, when associated withneurofibromatosis, to behave in an indolent mannerhas been noted previously."-'5 The overall incidenceof neurofibromatosis in our patients was 48%.This is comparable to that observed in some otherreports,'2" but rates in other series have been lower(10}30%). 415 16

We have shown in this series that, while half ofthe optic nerve gliomas remained quiescent, theremainder enlarged. The crucial question is whetherthis enlargement is due to division of tumour astro-cytes with spread along the optic nerve or is merely anincrease in bulk caused by the accumulation ofmucin or an associated desmoplastic reaction in themeninges. While the increase in the bulk of thetumour can be easily demonstrated with CT or evenbetter with MRI,'7" it is extremely difficult to definethe extent of the astrocytoma. The histopathologistexamining appropriately stained sections of tissueobtained from the edge of a glioma will often find itdifficult to distinguish between abnormal and normalastrocytes. Abnormal astrocytes can be present in theoptic nerve or chiasm without significant enlarge-ment, as evidenced in our patients, where the chiasmlooked normal but there was tumour at the site ofexcision. The most sophisticated imaging techniquescurrently available are therefore incapable of definingthe true edge of a glioma.

971


http://bjo.bmj.com

/B

r J Ophthalm


ber 1989. Dow

nloaded from

http://bjo.bmj.com/


Physiological tests can be used to determine theextent of lesions within the optic pathways.2" Visualevoked potentials produced by stimuli in differentparts of the visual fields or careful evaluation of thevisual fields will demonstrate functional defects pro-duced by tumour. However, quite extensive gliomasoften cause only minimal compromise of conduction,since these tests, like the imaging techniques, under-estimate the extent of a glioma. This was exemplifiedby the three patients with an active glioma wheretumour was found in the cut end of the optic nerveand assiduous examination of the contralateral visualfield and VEP studies failed to demonstrate anyabnormality of chiasmal function, though onedeveloped a field defect after 24 months. Lewis et al"CT scanned 217 patients with neurofibromatosis andfound tumours of the anterior visual pathways in15%. Two-thirds of these tumours were entirelyunsuspected. Similarly Spencer' reported cases ofbilateral thickening of the optic nerves when clinicallyonly one eye was affected. Two of our patients had asimilar abnormality.The difficulties in determining the natural history of

these lesions was only partly due to our inabilityto assess their extent accurately. An even moreimportant factor was our decision about eight yearsafter the start of our study to excise the morerapidly growing and presumably more aggressivetumours.During the early years of this study all patients

were managed as though the gliomas would behavelike a hamartoma. All 16 patients in the quiescentgroup have behaved in this manner, and none hasdeveloped evidence of invasion of the chiasm duringquite long period of follow-up. Two of the first fivepatients with a more active course developed acontralateral field defect. One of these patients diedduring a pneumoencephalogram which was beingcarried out to determine the size of the chiasm. Theseevents profoundly altered our attitude to the best wayof dealing with these tumours. Subsequently addi-tional evidence was provided by one of the patients inwhom the optic nerve was excised, with microscopictumour at the proximal cut end. This patient thendeveloped a progressive field defect and enlargementof the contralateral optic nerve. We believe thatthese cases show that an optic nerve glioma does havethe potential to expand within or extend into thechiasm in a similar manner to the case reported byMcDonnell and Miller"2: an 11-year-old boy with anoptic nerve glioma initially confined to the orbit andoptic canal developed neuroradiological evidence ofintracranial extension to involve the hypothalamusand third ventrical.The results of surgical intervention have been

instructive. One patient of the 10 who had a cranio-

tomy had an enlarged chiasm, and the optic nervewas not excised. Three of the remainder had tumourat the proximal cut end of the nerve, making itprobable that the chiasm was infiltrated by tumoureven though it appeared normal to the surgeon.MRI, which was not used in any of these patients,may be capable of detecting small areas of involve-ment better than a surgeon using an operatingmicroscope. None of our patients with normal MRIstudies have had histological evidence of tumour inthe chiasm, and two patients who had an enlargedchiasm with normal fields and CT scans weredetected with MRI. It will be interesting to see towhat extent MRI will enable us to select only thosepatients with an uninvolved chiasm.

There has been no evidence of tumour growth ineither the chiasm or the contralateral optic nerve inany of the six patients where the glioma was totallyexcised during a follow-up period ranging from 2-10years. In the three patients who had tumour up to theedge of excision, only one has shown an increase inthe size of the chiasm with the development of anenlarged contralateral optic nerve associated withdisc oedema and an abnormal field and VEP. Thecentral vision has remained stable at 6/6 over sevenyears. The other two patients have shown no evidenceof spread of their residual tumour over periods of sixand nine years. This is in direct contrast to the activityand enlargement of the orbital portion of the tumourprior to craniotomy. Others have noted that incom-plete removal of juvenile optic nerve gliomas doesnot necessarily lead to later clinical recurrence,and it has been suggested that removal of the mainpart of the tumour might remove the stimulus to itsperipheral growth.'6 It could be argued that evenpatients with evidence of chiasmal involvement mightbenefit from removal of the optic nerve glioma, butwe believe this approach is unjustifiable withoutmore definite evidence. Further detailed follow-up ofthese patients will hopefully provide more evidenceabout the natural history and effect of treatment onthis group of patients.

In reviewing our patients over the last 15 years wehave concluded that, while some gliomas are relativ-ely inactive and may well have undergone thedeceleration in growth which Alvord et al havesuggested, others show a higher level of biologicalactivity, with enlargement and worsening of theclinical picture. One cannot equate this growth oftumour with spread along the optic pathway, butthere is sufficient evidence from our study and that ofothers to indicate that these tumours do pose athreat by direct extension of the astrocytoma into apreviously uninvolved chiasm. This threat should beremoved by total excision of the involved optic nerveparticularly where vision is poor. We do not believe

972


http://bjo.bmj.com

/B

r J Ophthalm


ber 1989. Dow

nloaded from

http://bjo.bmj.com/

Optic nerve glioma and the management ofoptic nerve tumours in the young

Management of Primary Optic NerveTumours in the Young

Full clinical andGood Visions radioical wo rku

Poor Vision+_ ~~~~~adiological workup

MRI, CT etc MRI, CT etcsuggest glioma suggest PONM

Progression or change ofclinical diagnosis to PONM

v

Chiasm macroscopically Chiasm macroscopically

flt ~~~~normal involved with gliomaNo progression 'All"

Visiolgood Vision poor

Large biopsy|(frozen section)

Glioma Meninaioma

v MenlInglI

+ ?Radiotherapy

Observe Excise whole N ?Surgical debull

Fig. 4 Flow-chart ofrecommended management ofoptic nerve tumours in young patients.

that clinicians who have responsibility for thesepatients should adopt a totally passive role.

The management ofprimary optic nerve tumours inyoung patients without clinical, physiological, or

radiological evidence ofchiasmal involvementIn another communication' we have drawn attentionto the aggressive nature of primary optic nervemeningiomas in young patients. These tumours arecapable of invading the middle fossa quite quickly. Itis important, therefore, that all primary optic nervemeningiomas which occur under the age of 30 shouldbe diagnosed early and a neurosurgical clearance ofthe tumour carried out via a transfrontal cranio-orbitotomy. In the present communication we haveshown that a significant proportion of optic nervegliomas behave in an active manner and can spreadfrom the nerve to the chiasm and affect the vision inthe other eye. Their excision appears to remove thisthreat. In young patients it is therefore logical toexcise all primary optic nerve meningiomas and theactive group of optic nerve gliomas, preserving the

loma(extensive)

Attempt removal

(if feasible)

optic nerve in patients with an indolent or inactiveglioma.

In most patients the correct identification of aninactive glioma is not difficult and is based on thetypical clinical and radiological appearances.27 How-ever, a few patients have features which make thedifferentiation of meningioma from glioma verydifficult even with MRI. It is always possible that thetumour is a meningioma which can spread aggress-ively (meningioma cases 1 and 28). This danger can bereduced by excising all optic nerve tumours via atransfrontal approach where vision is poor, a proce-dure we think justifiable on the grounds that we haveseen almost as many patients under the age of 30 withmeningioma'3 as inactive glioma.'6 The pathologistcan then decide on the nature of the tumour.

Patients with good vision (6/12 or better)If MRI and CT scans show an appearance compatiblewith primary optic nerve meningioma then thepatient should be referred for transfrontal neuro-surgical exploration. A biopsy should be sent for

V

973


http://bjo.bmj.com

/B

r J Ophthalm


ber 1989. Dow

nloaded from

http://bjo.bmj.com/


frozen section, and if the diagnosis of meningioma isconfirmed then the optic nerve should be completelyexcised together with the tumour. If it is glioma, theoptic nerve should be preserved.The remainder are patients with either a typical

glioma or an optic nerve tumour of doubtful nature.These should be observed to see if there is any changein the size of the tumour, the vision, or the character-istics of the MRI and CT scans which would suggestmeningioma or active glioma.

Patients with poor vision (6118 or worse)These patients should have a transfrontal explorationof the chiasm. If despite the normal investigations thechiasm is involved by glioma, further surgery is notindicated. In all other patients the optic nerve shouldbe excised together with the tumour from justanterior to the chiasm to the posterior surface of theglobe. If meningioma has spread into the middlefossa, the neurosurgeon will carry out a microsurgicalremoval of the tumour, if feasible, and possibly treatthe area with postoperative radiotherapy. In occas-

sional patients in whom the history suggests thetumour has been present unchanging for a number ofyears it may be reasonable to continue to observe thepatient.The accompanying flow chart (Fig. 4) indicates the

salient points in our management of primary opticnerve tumours in young patients.

Our thanks are due to all the clinicians who referred these patients,Mr Alan Crockard and Professor Lindsay Symon, neurosurgeons atthe National Hospitals, Queen Square and Maida Vale, Dr G A SLloyd and Dr Ivan Moseley, Radiology Department, the staff of theMedical Illustration Department, and Mrs S J Cole, who helped withthe typing of the manuscript.

AAMcN was supported by the Royal Australian College ofOphthalmolgists-OPSM travelling fellowship.

References

I Hoyt WF, Baghdassarian SA. Optic glioma of childhood.Natural history and rationale for conservative management.Br J Opthalmol 1969; 53: 793-8.

2 Glaser JS, Hoyt WF. Corbett J. Visual morbidity with chiasmalglioma. Arch Ophthalmol 1971; 85: 3-12.

3 Spencer WH. Primary neoplasms of the optic nerve and itssheath; clinical features and current concepts of pathogeneticmechanisms. Trans Am Ophthalmol Soc 1972; 70: 490-528.

4 Wong IG, Lubow M. Management of optic glioma of childhood:a review of 42 cases. Neuro-ophthalmology. St Louis: Mosby,1972:6:51-60.

S Wright JE, McDonald WI, Call NB. Management of optic nerve

gliomas. Br J Ophthalmol 1980; 64: 545-52.

6 Wong JYC, Uhl V, Wara WM, Sheline GE. Optic gliomas. Areanalysis of the University of California experience. Cancer1987; 60: 1847-55.

7 Alvord EC, Lofton S. Gliomas of the optic nerve or chiasm.Outcome by patients' age, tumor site, and treatment.J Neurosurg 1988; 68: 85-98.

8 Wright JE, McNab AA, McDonald WI. Primary optic nervesheath meningioma. Br J Ophthalmol 1989; 73: 960-6.

9 Hoyt WF, Meshel LG, Lessell S, Schatz NJ, Suckling RD.Malignant optic glioma of adulthood. Brain 1973; 96: 121-32.

10 Crowe FW, Schull WJ. Diagnostic importance of cafe-au-laitspot in neurofibromatosis. Arch Intern Med 1953; 91: 758-66.

11 Martin P. Cushing H. Primary gliomas of the chiasm and theoptic nerves in their intracranial portion. Arch Ophthalmol 1923;52; 209-41.

12 Lloyd LA. Gliomas of the optic nerve and chiasm in childhood.Trans Am Ophthalmol Soc 1973; 71: 488-535.

13 Klug GL. Gliomas of the optic nerve and chiasm in children.Aust NZ J Surg 1977; 47: 596-6(X).

14 Fletcher WF, Imes RK, Hoyt WF. Chiasmal gliomas: appearanceand long-term changes demonstrated by computerized tomo-graphy. J Neurosurg 1986; 65: 154-9.

15 Yanoff M, Davis RL, Zimmerman LE. Juvenile pilocyticastrocytoma ('glioma') of the optic nerve. Clinicopathologicalstudy of sixty-three cases. In: Jakobiec FA, ed. Ocular andadne.ral tumors. Birmingham: Aesculapius, 1978: 685-7(07.

16 Rush JA, Younge BR. Campbell RJ, MacCarty CS. Opticglioma. Long-term follow-up of 85 histopathologically verifiedcases. Ophthalmology' 1982; 89: 1213-9.

17 Holman RE, Grimson BS, Drayer DP, Buckley EG, BrennanMW. Magnetic resonance imaging in the evaluation of opticgliomas. Am J Ophthalmol 1985; 100: 596-60(1.

18 Haik BG, Saint Louis L, Bierly J, et al. Magnetic resonanceimaging in the evaluation of optic nerve gliomas. Ophthalnology1987; 94:709-17.

19 Eidelberg D, Newton MR, Johnson G, it al. Chronic ulilateraloptic neuropathy: a magnetic resonance study. Ann Neurol 1988;24:3-11.

20 Grosswasser Z, Kriss A, Halliday AM, McDonald WI. Patternand flash evoked potentials in the assessment and management ofoptic nerve gliomas. J Neurol Neurosurg Psychiatr~y 1985; 48:124-7.

21 Lewis RA, Gerson LP, Axelson KA, Riccardi VM, WhitfordRP. von Recklinghausen's neurofibromatosis 11. Incidence ofoptic gliomata. Ophthalniology 1984; 91: 929-35.

22 McDonnell P. Miller NR. Chiasmatic and hypothalamic exten-sion of optic nerve glioma. Arch Ophthalmol 1983; 101: 1412-5.

23 Hudson AC. Primary tumours of the optic nerve. R LondOphthalmic Hosp Rev, 1912; 18: 317-439.

24 Marquanrdt MI), Zimmerman LE. Histopathology ofmeningiomas and gliomas of the optic nerve. Hurn Pathol 1982;13: 226-35.

25 Borit A, Richardson EP. The biological and clinical behaviourof pilocytic astrocytomas of the optic pathways. Brain 1982; 105:161-87.

26 Anderson DR, Spencer WH. Ultrastructural and histochemicalobservations of optic nerve gliomas. Arch Ophthalmol 197(0; 83:324-35.

27 Jacobiec FA, Depot MJ, Kennerdell JS, et al. Combinedclinical and computed topographic diagnosis of orbital gliomaand meningioma. Ophthalmnology 1984; 91: 137-55.

Accepted for publication 25 May 1989.

974


http://bjo.bmj.com

/B

r J Ophthalm


ber 1989. Dow

nloaded from

http://bjo.bmj.com/

optic nerve glioma and the of optic nerve tumours...

Documents