Download pdf - Primary intraorbital extraocular primitive neuroectodermal ...bjo.bmj.com/content/bjophthalmol/70/7/543.full.pdf · neuroectodermal(neuroepithelial) tumour ... Theneurogliawasidentified

British Journal of Ophthalmology, 1986, 70, 543-548

Primary intraorbital extraocular primitiveneuroectodermal (neuroepithelial) tumourSAMRUAY SHUANGSHOTI,' WIT MENAKANIT,2 WITIT CHANGWAIVIT,3 ANDNITAYA SUWANWELA4

From the 'Department of Pathology, Faculty of Medicine, Chulalongkorn University, Bangkok; the2Department of Pathology, Faculty of Medicine, Chiang Mai University, Chiang Mai; the 'Department ofRadiology, Faculty of Medicine, Chiang Mai University, Chiang Mai; and the 4Department of Radiology,Faculty ofMedicine, Chulalongkorn University, Bangkok, Thailand

SUMMARY A case is reported of primary primitive neuroectodermal (neuroepithelial) tumouroccurring in the right orbit of a 52-year-old man. The intraorbital extraocular location is unique forthis kind of neoplasm. The malignant tumour was differentiated into primitive neuroepithelial,ependymal, and oligodendroglial cells. The neuroglia was identified by localisation of cytoplasmicglial fibrillary acidic protein. It is suggested that this primary intraorbital, extraocular, primitiveneuroectodermal tumour with neuroglial differentiation is akin to the primitive neuroectodermaltumours of the neuraxis, including the cerebellar medulloblastomas, and to some peripheral nervetumours known as malignant neuroepitheliomas, malignant ependymomas, and neuroblastomas.The ectomesenchymal remnant of the neural crest or ectopic neuroepithelium or both may havebeen the origin of the present tumour. Differentiation of the neuroectodermal component of theneural crest or heterotopic neuroepithelium or both would create a primitive neuroectodermaltumour with diverse neuroglial elements.

Primitive neuroectodermal tumours of the neuraxis '-3and peripheral nerves45 have been described. How-ever, we are unable to trace a report of the same typeof tumour arising primarily within the orbit butoutside the eyeball. In this communication we de-scribe a primary intraorbital, extraocular, primitiveneuroectodermal tumour occurring in an adult man.

Case report

A 52-year-old man was admitted to hospital becauseof progressive right exophthalmos for one month. Hedenied ocular pain, blurred vision, diplopia, andother systemic symptoms.

Physical examination revealed a body temperatureof 370C, pulse rate 85 beats/min, respiratory rate20/min, and blood pressure 150/90 mmHg. The righteye showed mild oedema of the eyelids and prop-tosis. Each pupil, 3 mm in diameter, was reactive tolight. Ocular movements were bilaterally full in alldirections. The fundi showed constriction of the

Correspondence to Professor Samruay Shuangshoti, Department ofPathology, Chulalongkorn Hospital, Bangkok 10500, Thailand.

arterioles. No papilloedema or intraocular tumourwas observed. Cranial nerves, visual field, and visualacuity were normal, as were other systems of thebody and various laboratory tests. No abnormalmasses were detected in any part of his body.Computerised tomographic (CT) scans of the

orbits and brain disclosed a right enlarging intra-orbital tumour located laterally. The right eyeballand optic nerve were displaced medially (Figs. 1, 2).No intracranial lesions were seen.The patient underwent Kronlein's operation three

weeks after admission to hospital. The outer wall ofthe right orbit was resected. The laterally situatedintraorbital extraocular tumour was found to beunencapsulated and attached to the lateral rectusmuscle. There was no bone invasion. The lesionwas removed as much as possible. No connectionbetween the tumour and any nerve was observed. Itappeared clinically to be a neoplasm of the lacrimalgland.

PATHOLOGICAL EXAMINATIONThe formalin-fixed tissue fragment was white, firm,and 3-5 x 4-0 x 4-0 cm. It was embedded in paraffin

543

on 5 May 2018 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.70.7.543 on 1 July 1986. D

ownloaded from

http://bjo.bmj.com/

Samruay Shuangshoti, Wit Menakanit, Witit Changwaivit, and Nitaya Suwanwela

Fig. 2 Axial cut at the level ofthe optic nerves demonstratesa mass invading into the right lateral rectus muscle (thickarrow). Note also medial displacement ofthe right opticnerve (thin arrow) and mildproptosis ofthe right eyeball.

and sections were stained with haematoxylin andeosin (H-E). The following special stains weredone: Mallory's phosphotungstic acid haematoxylin(PTAH) for neuroglial fibres, Gomori's stain forreticulin fibres, and Nissl's stain for neuronal tigroidsubstance.The peroxidase-antiperoxidase technique was ap-

Fig. l A, B Axialand coronal CTscans oftheorbitsshow plied for immunohistochemical localisation of glialan infiltrative soft tissue mass in the superolateralpart ofthe fibrillary acidic protein (GFAP) in formalin-fixed andright orbit (left arrows) with pressure effecton the lateral wall paraffin-embedded tissue sections. A fragment of the(rightarrows). normal brain was similarly processed as a control.

Fig. 3 Microscopicfeatures oftumour. A: Numerous smallneoplastic cells in lobules and in nodefinite pattern ofarrangement areshown (H-E, x 32). B: HomerWright's rosettes areformed byprimitive neuroepithelial tumourcells. Note tangles ofcellularprocesses within the centre ofrosettes (H-E, x65)..

544



j.com/

Br J O


ownloaded from

http://bjo.bmj.com/

Primary intraorbital extraocularprimitive neuroectodermal (neuroepithelial) tumour

Microscopically, plentiful neoplastic cells of diver-gent types were found in lobules and in randomdistribution (Fig. 3A). Many small ovoid or carrot-shaped tumour cells with hyperchromatic nucleirimmed by scant cytoplasm and hair-like cytoplasmicpolar processes were often arrayed in rosettes thatcontained central tangles of these processes, repre-senting Homer Wright's rosettes (Fig. 3B). Theseneoplastic cells were regarded as primitive (undiffer-entiated) neuroepithelial cells which resembled bothneuroblasts and medulloblasts.

Several epithelium-like tumour cells had round toovoid vesicular nuclei which were rimmed by homo-geneous cytoplasm. Some had blue processes inPTAH preparations. They were occasionally arrangedin true rosettes with central canals (Fig. 4A) andperivascular pseudorosettes characterised by attach-ment of the cellular processes to the vascular walls(Fig. 4B). These two patterns of the neoplasm wereregarded as formed by ependymal cells.A moderate number of the tumour cells had well

defined outlines, distinctly clear perikaryon, andsmall dark nuclei giving a fried-egg appearance. Theywere interpreted as oligodendroglia (Figs. 4B, 5A).Neurons were not detected in Nissl's preparations.A small number of widely scattered rounded

tumour cells contained cytoplasmic GFAP and wereregarded as neuroglia. Some had small dark nucleisurrounded by moderate amount of cytoplasm; theywere interpreted as oligodendroglia (Fig. SB).Others had large hyperchromatic nuclei rimmedby scant perikaryon; they were regarded as undiffer-

entiated neuroglia (Fig. SC). Similar GFAP-positiveneuroglia, particularly astrocytes, were also seen inthe fragment of normal brain serving as a control.

Reticulin fibres were few in some parts of thetumour but were numerous elsewhere (Fig. 6). Thiswas also true for collagen fibres.The pathological diagnosis was primary intra-

orbital, extraocular, primitive neuroectodermaltumour showing neuroglial differentiation and pro-ducing right proptosis.

POSTOPERATIVE COURSEThe patient received a course of radiotherapy total-ling 5200 rad to the operative field over 11/2 months.At the end of the radiotherapeutic course his visionwas normal, as it was before the surgical intervention.The proptosis had improved, and the general con-dition was satisfactory. He was well when he was seenagain 6 months later.

Discussion

The primitive neuroectodermal tumour is defined asa malignant neoplasm arising from primitive (stemcell) neuroepithelium which is capable of differen-tiating into various derivatives within the neuronaland neuroglial lines. Pathologically, Homer Wright'srosettes formed by primitive neuroepithelial cells arepresent. True rosettes and perivascular pseudo-rosettes may be present. The tumour may show aprominent mesenchymal component. ' I

Fig. 4 Ependymal andoligodendroglial differentiation oftumour. A: A true rosetteformedby radially arranged ependymaltumour cells round central canal isshown (H-E, x 253). B: Aperivascularpseudorosetteconsisting ofcentral blood vesselwith processes ofependymaltumour cells attached to its wall isshown. Some rounded neoplasticcells having distinctly clearperikaryon roundsmall nuclei togivefried-egg appearance, as in acluster at the right lower corner ofthephotomicrograph, areconsidered to be oligodendroglia(H-E, x 63).

545



j.com/

Br J O


ownloaded from

http://bjo.bmj.com/


Fig.5 Oligodendroglial differentiation and resultofimmunohistochemicalstudy. A: Many oligodendroglia with fried-eggappearance are shown (H-E, x 100). B: A GFA P-positive tumour cell interpreted as oligodendrocyte (arrow) isdemonstrated (GFA Pstain, x 1000). C: A neoplastic cell having large hyperchromatic nucleus surrounded by scantGFA P-positive perikaryon regarded as undifferentiated neuroglial cell (arrow) is present among other undifferentiatedtumourcells (GFAPstain, x 1000).

The intraorbital extraocular neoplasm describedhere fulfils these criteria for designating it as aprimitive neuroectodermal tumour because it hadprimitive neuroepithelial cells forming HomerWright's rosettes, ependymal cells forming truerosettes and perivascular pseudorosettes, and neo-plastic oligodendroglia. The present of neurogliawithin the tumour was confirmed by the localisationof cytoplasmic GFAP in some tumour cells, as wehave demonstrated previously.`' In those studies allkinds of neuroglia (astrocytes, ependymal cells,oligodendroglia, and undifferentiated neuroglia)were GFAP-positive.'-"

Nevertheless, the presence of primitive neuro-

epithelial cells forming Homer Wright's rosettes inthe present tumour suggests that it is a neuro-blastoma. Indeed, neuroblastomas in adult patientshave been reported."' However, the concept of theprimitive neuroectodermal tumour encompasses aswell the neuroblastomas. A neuroblastoma can beperceived as primitive neuroectodermal tumour withentire neuroblastic differentiation. The present intra-orbital extraocular neoplasm is a primitive neuro-ectodermal tumour with neuroglial differentiation.Nevertheless a small number of GFAP-positiveneuroglial neoplastic cells as well as prominentreticulin and collagen fibres in some parts of thistumour may not support the diagnosis of the primitive

546



j.com/

Br J O


ownloaded from

http://bjo.bmj.com/

Primary intraorbital extraocularprimitive neuroectodermal (neuroepithelial) tumour

Fig. 6 Reticulinfibres withintumour. A: Few reticulinfibres arepresent aroundsome lobules ofthetumour cells but are absent withinlobules (Gomori's stain, x 72). B:Dense reticulinfibres areshown in afield ofless numerous neoplasticcells (Gomori's stain, x 72).

neuroectodermal tumour but favour a mesenchymalorigin of the neoplasm. According to our experience,however, ependymal and oligodendroglial tumourcells are not frequently GFAP-positive, as are astro-cytes, which were not seen in the present tumour. Aprominent mesenchymal component is known inprimitive neuroectodermal neoplasms.'We regard the current primary intraorbital, extra-

ocular, primitive neuroectodermal tumour as akin tothose neoplasms reported as neuraxial primitiveneuroectodermal tumours, including the cerebellarmedulloblastomas' " and to some peripheral nerveneoplasms which have been described as malignantneuroepitheliomas, '- malignant ependymoma,'9and primitive neuroectodermal tumour (neuro-blastoma). These peripheral nerve tumours con-tained Homer Wright's rosettes and true rosettes asin our patient's tumour relating to the orbit. Theyoften metastasised widely.5 1-The primitive (stem cell) neuroepithelium has

been suggested as the origin of the neuraxial primi-tive neuroctodermal tumours.3" For the peripheralprimitive neuroectodermal tumours an origin fromthe neural crest has been proposed.9 Evidence hasbeen presented relating to the combined mesen-chymal and neuroectodermal (ectomesenchymal)nature of the neural crest,' `- because it is formed bymingling of the mesenchyme and developing neuro-epithelium of the neural folds before closure of theneural tube.'7 The neural crest yields diverse deri-vatives of both mesenchymal and neuroepitheliallines which migrate throughout the body, particularlyin association with the peripheral nerves.'9 Althoughno connection of the present tumour with any majornerve was observed clinically, its origin from a smallnerve, such as a nerve twig, cannot be excluded.Moreover, in the development of the eyeball, whichhas been reviewed previously,2 the primitive neuro-

epithelium forming the optic primordium maybecome ectopic within the orbit, a comparabledevelopmental anomaly known as heterotopia of theneuraxis.' This heterotopic neural tissue and/orneuroectodermal component of the neural crestcomprising the sheath of the peripheral nerve withinthe right orbit may be the origin of our patientstumour, which further differentiated into diverseneuroepithelial derivatives.

Professor Samruay Shuangshoti is in receipt of funds from theAnandhamahidol Foundation, Medical Section, Bangkok, Thailand.

References

1 Hart MN, Earle K. Primitive neuroectodermal tumours of thebrain in children. Cancer 1973; 32: 890-7.

2 Kosnick EJ, Boescl CP, Bay J, Sayers MP. Primitive neuro-ectodermal tumours of the central nervous system in children.J Neurosurg 1978; 48: 741-6.

3 Rorke LB. The cerebellar mcdulloblastoma and its relationshipto primitive neuroectodermal tumours. J Neuropathol ExpNeurol 1983;42: 1-15.

4 Seemayer TA, Thelmo WL, Bolande RP, Wiglesworth FW.Peripheral ncurocctodermal tumours. Per.spect Pediatr Pathol1975; 2: 151 -72.

5 Nesbitt KA, Vidone RA. Primitive neuroectodermal tumour(neuroblastoma) arising in sciatic nerve of a child. Cancer 1976;37: 1562-70.

6 Shuangshoti S. O'Charocn S. Cerebellar neoplasm of mixedmesenchymal and neuroepithelial origin: case report. J Neuro-surg 1983; 59: 337-43.

7 Shuangshoti S. Suwanwela N, Suwanwela C. Combined occur-rencc of paraganglioma and glioma of cons mcdullaris andcauda cquina. J Surg Oncol 1984; 25: 162-7.

8 Shuangshoti S, Kasantikul V, Suwanwela N, Suwanwela C.Solitary intracranial cxtraccrebral gliomas: case report.J Neurosurg 1984; 61: 777- 1.

9 Shuangshoti S. Kasantikul V, Suwangool P. Chittmittrapap S.Malignant neoplasm of mixed mesenchymal and neuroepithelialorigin (ectomesenchymoma) of thigh. J Surg Oncol 1984; 27:208 13.

It) Mackay B. Luna MA, Butler JJ. Adult neuroblastoma: electronmicroscopic observations in nine cases. Cancer 1976; 37:1334-51.

547



j.com/

Br J O


ownloaded from

http://bjo.bmj.com/


11 Palmer JO, Kasselherg AG, Netsky MG. Differentiation ofmedulloblastoma: studies including immunohistochemical local-isation of glial fibrillary acidic protein. J Neurosurg 1981; 55:161-9.

12 Stout AP. A tumour of the ulnar nerve. Proc N Y Pathol Soc1918; 18:2-11.

13 Lagerkvist B, Ivemark B, Sylven B. Malignant neuroepi-thelioma in children: a report of three cases. Acta Chir Scand1969; 135:641-5.

14 Bolen JW, Thorning D. Peripheral neuroepithelioma: A lightand electron microscopic study. Cancer 1980; 46: 2456-62.

15 Lanford JA, Cohen I. Ependymal neoplasm of median nerve

with case report. South MedJ 1927; 20: 273-8.16 Shuangshoti S. Chongehet V. Malignant mesenchymoma of

ulnar nerve: combined sarcoma of nerve sheath and rhabdomyo-sarcoma. J Neurol Neurosurg Psychiatry 1979; 42: 524-8.

17 Shuangshoti S. Melanotic mucin-producing neuroepithelial neo-

plasm of mesencephalon with consideration of similar tumours indifferent locations. J Neurol Neurosurg Psychiatry 198t); 43:810-7.

18 Shuangshoti S, Benjavongkulchai S, Chittmittrapap S. Malig-nant mesenchymoma of median nerve: combined nerve sheathsarcoma and liposarcoma. J Surg Oncol 1884; 25: 119-23.

19 Weston TA. The migration and differentiation of neural crestcell. In: Abercrombie M, Brachet J, King TJ. eds. Advances inmorphogenesis. New York: Academic Press, 197t): 41- 114.

20 Shuangshoti S. Meningioma of the optic nerve. BrJ Ophthalmol1973;57:265-9.

21 Shuangshoti S, Yenrudi S. Netsky MG. Heterotopias ofrain as

a cause of infantile regurgitation. Teratology, 1981; 23: 67-73.

Acceptedfor publication 21 Novemnber 1985.

548



j.com/

Br J O


ownloaded from

http://bjo.bmj.com/