Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Editor,—Recently, Margo and coworkers1described a case of intraepithelial sebaceous carcinoma of the conjunctiva and eyelid erroneously diagnosed as a unilateral blepharoconjunctivitis for 10 years. This was reported to be the first published case of de novo sebaceous carcinoma in the conjunctival epithelium without evidence of a tumour focus within the eyelid. We present another case of diffuse intraepithelial sebaceous carcinoma of the conjunctiva presenting as blepharoconjunctivitis and superior limbic keratitis of 10 months’ duration.
A 70-year-old woman presented with a 10 month history of left sided, unilateral redness, tearing, and swelling of the lids. Examination showed marked conjunctival injection and thickening along the superior temporal and inferotemporal limbus, with moderate lid thickening, and pale corneal changes (Fig 1). There was mild foreshortening and scarring of the inferior fornix. A presumptive diagnosis of carcinoma in situ was considered with a secondary diagnosis of sebaceous gland carcinoma with intraepithelial spread. Biopsies of several sites of limbal, bulbar, and palpebral conjunctiva (Fig 2) revealed a diffuse intra-epithelial malignancy with pagetoid and bowenoid growth and moderate chronic non-granulomatous inflammation in the underlying substantia propria but no evidence of stromal invasion. The tumour cells were markedly polymorphic and often exhibited multiple nucleoli. Mitotic figures were frequent (Fig 2B). Occasional dyskeratosis was noted but in many tumour cells the cytoplasm appeared foamy. Stains for mucin, mucopolysaccharides, and melanin within tumour cells were inconclusive, nor could lipid be demonstrated in either frozen or paraffin sections. Staining with OM1 antibody (kindly provided by Dr H Grossniklaus, Emory University, Atlanta, GA, USA) against sebaceous gland antigen, however, revealed positive labelling of tumour cells.
Exenteration was performed and histology showed diffuse infiltration of the tarsal and bulbar conjunctiva by tumour cells, extending intraepithelially across the superior, and to a lesser degree the inferior, limbus onto the cornea. The inferior tarsal plate was largely destroyed and replaced by irregular fibrotic scar tissue, within which a small focus of tumour cells in a sebaceous gland was observed. At the inferior medial lid margin intraepithelial tumour spread into the punctum lacrimale and upper canaliculus was present but the lacrimal sac was free of malignant infiltration. In this area, another focus of intraepithelial malignancy was detected at the base of a pilosebaceous follicle while the overlying surface epithelium was apparently normal.
The origin of the intraepithelial form of sebaceous carcinoma of the eyelid is still uncertain.1-4 While some believe that sebaceous metaplasia of the epithelium must precede neoplastic transformation, Margo et al 1 have suggested that the tumour may arise de novo within the conjunctival epithelium since, in their case, no clear focus of neoplasia could be demonstrated in sebaceous glands of the eyelid.
In the present case, serial sections revealed evidence of a connection of the intraepithelial changes to a focus of neoplasia in a sebaceous gland of the lower lid. However, the surrounding basal lamina of this sebaceous gland remained intact. In addition, invasion of the basal lamina could not be demonstrated in any of the multiple biopsies or in the exenteration specimen. This pattern of spread is of particular interest as extensive intraepithelial neoplasia of the conjunctiva and cornea was most prevalent superiorly, and the meibomian glands of the upper lid appeared unremarkable. Rao et al 2have suggested that intraepithelial spread from a nidus of sebaceous tumour demonstrates decreasing epithelial involvement furthest from the tumour. This case appears to contradict the logic of this pattern of spread.
These findings, however, do not exclude the possibility of multicentric origin of the tumour. The additional focus of carcinomatous change in a pilosebaceous unit of the caruncle is suggestive of multicentricity as serial sections revealed no contiguity with surface intraepithelial neoplasia. This may represent separate origins of carcinoma as a result of some innate cellular abnormality of epithelium in this patient.
Intraepithelial spread of sebaceous carcinoma is associated with a poor prognosis.5 Despite the absence of invasion in this case, exenteration was considered the appropriate therapy since a localised conjunctival and corneal epithelial excision in this case was not feasible. The presence of a tumour nidus in the lower lid suggests that this latter course of action would not have proved definitive. The presence of extension into the lacrimal punctum lends further credence to this hypothesis.