British,Journal of Ophthalmology 1992; 76: 634-636

Myoepithelioma of the lacrimal gland: report ofacase with spindle cell morphology

Ramon L Font, Alec Garner

AbstractThe case is described of a 23-year-old femalepatient presenting with unilateral proptosis,headaches, and transient epiphora. Surgeryrevealed an encapsulated tumour composedexclusively of spindle-shaped cells within arichly vascularised myxoid stroma. Immuno-histochemical staining showed focal positivityfor smooth muscle actin, vimentin, and glialfibriliary acidic protein. These combinedfindings are interpreted as providing evidenceof a myoepithelioma, which may be regardedas a monomorphic adenoma consisting solelyofmyoepithelial cells. To our knowledge this isonly the second report of such a tumour in thelacrimal gland.(BrJ7 Ophthalmol 1992; 76: 634-636)

Itmay be that most benign adenomatous tumoursof the lacrimal gland are derived from themyoepithelial component of the parenchyma.These cells differentiate to produce unmistake-able epithelial and, frequently, overtly glandularstructures as well as, in the vast majority ofinstances, a variably hyaline, myxoid, orchondroid extracellular stroma. The resultanttumour is then aptly named a pleomorphicadenoma, and in this respect the lacrimal glandparallels its salivary counterpart. Occasionalsalivary gland adenomas, however, are mono-

morphic in that fully differentiated ductular orduct-derived epithelium alone is seen, whileeven more unusually the tumours are restrictedto a myoepithelial proliferation with an absenceof manifest glandular components. Mono-morphic adenomas arising in the lacrimal glandare still rarer: tumours with a uniform glandularpattern have yet to be described and there hasbeen only one previous report of a pure myo-

epithelial proliferation.' Here we describe asecond case.

,. r~~~~~~~~~~~~~~~~~~~~~~~~~4

at r t f, ,f

.. .- "....l. .....

tumour (haematoxylin and eosin, x32).

sonography produced abnormal echoes temporal

to the globe, both superiorly and inferiorly.

The preoperative clinical diagnosis was

lacrimal gland tumour, probably benign, and 7

weeks later a circumscribed and apparently

encapsulated mass was excised through a left

lateral orbitotomy. There was no detectable

involvement of the adjacent periosteum and the

postoperative course was uneventful.

At the present time, more than 12 years after

surgery, the patient is alive and well with no sign

of recurrent tumour.

Pathology

Grossly, the resected tumour showed a smooth,

encapsulated, pale grey, firm and reniform mass

measuring approximately 30x25x 17 mm.

Microscopically, the mass was covered by a

fibrous capsule enclosing a solid tumour

composed exclusively of spindle-shaped cells.

Lobules of atrophic lacrimal gland tissue were

confined to one edge of the tumour adjacent to its

capsule (Fig 1). Most of the tumour cells had

rounded or elongated nuclei according to the

Department ofPathology, Institute ofOphthalmology, LondonA Garner

Department ofOphthalmology, CulienEye Institute, BaylorCollege of Medicine,Houston, TexasR L FontCorrespondence to:Professor A Garner, Instituteof Ophthalmology, 17-25Cayton Street, London EC1V9AT.

Accepted for publication9 April 1992

Case reportA 23-year-old woman presented to the owiUtai

clinic of Moorfields Eye Hospital complaining ofgradually increasing proptosis ofher left eye. Shehad been conscious of some 'puffiness' aroundthe eye for up to 1 year and for the last 3 monthshad suffered occasional left parietal headachesand transient epiphora. Ophthalmic examinationrevealed 5 mm proptosis with 2-3 mm down-ward displacement of the left eye and there are

some restriction of eye movement on upgaze,

particularly in abduction. Visual acuity was notaffected (RE 6/6; LE 6/4) and the ocular fundiwere ophthalmoscopically normal. A mass couldbe palpated in the left lacrimal fossa and ultra-

Figure 2 The lacrimal gland mass is covered by afibrouscapsule and consists ofspindle-shaped cells intermixed withnumerous capillaries (haematoxylin and eosin, x80).

634

on February 23, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.76.10.634 on 1 O

ctober 1992. Dow

nloaded from

Myoepithelioma ofthe lacrimal gland: report ofa case with spindle cell morphology

Figure 3 The centre ofthe tumour ts similarly composed of Figure 5 The capsule of the tumour seen in the upper part ofspindle-shaped cells with thin, elongated, cytoplasmic the photomicrograph, includes several lacrimal gland ducts,processes embedded in a myxoid matrix rich in capillaries while the underlying tumour shows cohesive perivascular cells(haematoxylin and eosin, x 80). alternating with loose areas consisting ofproteinaceous

exudate (Masson trichrome, x 80).

cellular orientation with inconspicuous nucleolibut occasionally slightly enlarged and hyper-chromatic nuclei were also seen. Mitotic figureswere rare. The cell cytoplasm was either eosino-philic or vacuolated, especially in less denselypopulated parts ofthe tumour where a prominentmyxoid stroma was observed (Figs 2 and 3). Arich capillary network with scattered largesinusoidal vessels was present. In some areas thetumour displayed confluent cystoid spacescontaining proteinaceous exudate and occasionalerythrocytes (Fig 4). The loose myxoid areasalternated with dense aggregates of spindle cells,some of which demonstrated a perivasculararrangement (Fig 5). Areas of necrosis withsecondary haemorrhage were present and inplaces this appeared to have been superseded byhyaline fibrous tissue.

Immunohistochemical procedures using aperoxidase-antiperoxidase method demonstratedfocal response of some cells with antibodiesagainst smooth muscle actin, vimentin, and glialfibrillary acidic protein (GFAP). Labelled anti-body to low molecular weight keratin (AE1)stained the residual lacrimal ducts adjacent to thecapsule of the tumour but the spindle cells wereunreactive (Fig 6). Negative staining reactionswere obtained using antibodies to S-100 protein.

DiscussionIt is a common finding that a proportion ofpleomorphic adenomas in both the salivary2 andthe lacrimal3 glands contain solid foci of myo-epithelial cell proliferation. The validity of a

Figure 4 In other areas the mass exhibits cystoid spacescontaining proteinaceous exudate and erythrocytes (Massontrichrome, x80).

diagnosis of myoepithelioma, however, dependson demonstrating that the tumour is essentiallymonomorphic and that the constituent cells aremyoepithelial. While this can reasonably beconstrued from the shape of the cells in thepresent case, the diagnosis is rendered moresecure by the evidence that a proportion of thespindle-shaped cells contained smooth muscleactin. Grossniklaus and colleagues,4 in animmunohistochemical study of normal andneoplastic lacrimal gland tissue, found that myo-epithelium in situ at the perimeter of normalsecretory acini reacts strongly with antisera tosmooth muscle actin, rarely with antibodies toGFAP and strongly but inconsistently with awide spectrum of antikeratin antibodies. Thesame study showed that a minority of thespindle-shaped stromal cells in pleomorphicadenomas were positive for smooth muscle actin,while rare stromal cells stained weakly forvimentin. In their study, positive reactions forkeratin and GFAP were associated with aproportion of so-called epithelioid as opposed tospindle-shaped myoepithelial cells. A weakresponse for GFAP was the sole positiveimmunohistochemical staining seen in the othercase of lacrimal gland myoepithelioma.' It wouldseem, however, that caution must be used in theinterpretation of such data. Thus, althoughGFAP reactivity has been used as a criterion forthe identification of myoepithelial cells, it waspositive in only two of 23 normal salivary glandsin one extensive study.5 Similarly, S-100 protein,

Figure 6 Section ofan area adjacent to that seen in Figure 5showing lacrimal gland ducts immunoreactive for cytokeratin(brown-black), in contrast to the underlying spindle cellswhich are negative (low molecular weight keratin AE I, x 80).

635

on February 23, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.76.10.634 on 1 O

ctober 1992. Dow

nloaded from

Font, Garner

the S-1OOb form of which has formerly beenregarded as a marker for myoepithelium,6 hasmore recently been shown to be confined toclosely associated autonomic nerves within theglands.7 That being the case, the negative reactionfor S-100 protein in both our patient and that ofHeathcote and colleagues' is not contrary to adiagnosis of myoepithelioma.

Ultrastructural characteristics of myo-epithelial cells, such as micropinocytotic vesicles,cytoplasmic filaments consistent with actin and apericellular basement membrane, have beendescribed in myoepithelial tumours of thesalivary glands8 and in the one recorded lacrimalgland example of this type of tumour.' Webelieve, however, that these are not patho-gnomonic features and are not essential for thediagnosis and note that Dardick and others8defined myoepitheliomas solely on the basis oftheir light microscopic and immunohisto-chemical properties. We contend, therefore, thatit is reasonable to regard the present tumour as aspindle-cell myoepithelioma and, despite thefinding of occasional cells with abnormal nuclei,to have anticipated a benign prognosis borne outby the patient's 12-year trouble-free post-operative survival.The potential for a variety of adenoma types,

ranging from pure ductal proliferations(monomorphic type) through lesions of mixedcell type to pure myoepithelial tumours, to beseen in the lacrimal gland is readily appreciatedon the assumption that, as has been reasoned inthe context of the salivary glands,5 they alloriginate from a common precursor. Whether ornot that precursor is myoepithelial, as hasgenerally been assumed, is not clear because ofinconsistent and potentially ambiguous immuno-histochemical marker findings. It is also reason-able to assume that most adenomas will occupythe mid-range of the spectrum and be mixedtumours, though it might be wondered why theextremes are encountered less frequently in thelacrimal gland than in the salivary glands. Using

the salivary tumours as a guide, however, there isno cause to regard monomorphic adenomas,including myoepitheliomas, as having a differentbiological behaviour from the more usualpleomorphic lesions.9 This is a point of someimportance because myoepitheliomas involvingthe salivary glands have quite often beendiagnosed as malignant tumours,'° as for a timewas the present case. As applied in the salivarygland context, a number of morphologicallydifferent subtypes of myoepithelioma aredescribed with 90% consisting of spindle-shapedor polygonal cells.8 We believe that bothHeathcote and colleagues' case and ours, areexamples of pure spindle cell myoepithelialtumours. Additionally, there is a third caserecorded by Rootman" which we suspect repre-sents the same entity.

We are grateful to Mr John Wright, Moorfields Eye Hospital, foraccess to the clinical information.

1 Heathcote JG, Hurwitz JJ, Dardick I. A spindle-cell myo-epithelioma of the lacrimal gland. Arch Ophthalmol 1990;108: 1135-9.

2 Thackray AC, Sobin LH. Histological typing of salivary glandtumours. Geneva: World Health Organisation, 1972; 9-28.

3 Ashton N. Epithelial tumours of the lacrimal gland. In:Bleeker GM, Garston JB, Kronenberg B, Lyle TK, eds.Orbital disorders. Basel: Karger, 1975; 306-23.

4 Grossniklaus HE, Abbuhl MF, Mclean IW. Immunohisto-logic properties of benign and malignant mixed tumor of thelacrimal gland. AmJ Ophthalmol 1990; 110: 540-9.

5 Stead RH, Qizilbash AH, Kontozoglou T, Daya AD, RiddellRH. An immunohistochemical study of pleomorphicadenomas of the salivary gland: glial fibrillary acidic protein-like immunoreactivity identifies a major myoepithelialcomponent. Hum Pathol 1988; 19: 32-40.

6 Hara K, Ito M, Takeuchi J, Iijima S, Endo T, Hidaka H.Distribution of S-100b protein in normal salivary glands andsalivary gland tumors. Virchows Arch[A] 1983; 401: 237-49.

7 Dardick I, Stratis M, Parks WR, DeNardi FG, Kahn HJ. S-100 protein antibodies do not label normal salivary glandepithelium. AmJ Ophthalmol 1991; 138: 619-28.

8 Dardick I, Thomas MJ, van Nostrand AWP. Myoepithelioma- new concepts of histology and classification: a light andelectron microscopic study. UltrastructPathol 1989; 13:187-224.

9 Cho KJ, Kim YI. Monomorphic adenomas of the salivaryglands: a clinicopathologic study of 12 cases with immuno-histochemical observation. Pathol Res Pract 1989; 184: 614-20.

10 Sciubba JJ, Brannon RB. Myoepithelioma of salivary glands:report of 23 cases. Cancer 1982; 49: 562-72.

11 Rootman J. Disease of the orbit: a multidisciplinaty approach.Philadelphia: Lippincott, 1988: 394-405.

636

on February 23, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.76.10.634 on 1 O

ctober 1992. Dow

nloaded from