SYMPOSIUM ON OPHTHALMIC CYTOLOGY
Year : 2007 | Volume
: 24 | Issue : 1 | Page : 22--26
Role of scrape cytology in the diagnosis of ocular surface squamous neoplasia
Tata Memorial Hospital, Mumbai, India
S V Kane
603, Yashowan Towers, 97, T.H.Kataria Marg, Behind Mahim Main Post Office, Mahim (West), Mumbai- 400016.
|How to cite this article:|
Kane S V. Role of scrape cytology in the diagnosis of ocular surface squamous neoplasia.J Cytol 2007;24:22-26
|How to cite this URL:|
Kane S V. Role of scrape cytology in the diagnosis of ocular surface squamous neoplasia. J Cytol [serial online] 2007 [cited 2020 Apr 2 ];24:22-26
Available from: http://www.jcytol.org/text.asp?2007/24/1/22/42086
The cornea, conjunctiva, and the limbus comprise the tissues at the ocular surface. The limbus is the most common site for occurrence of ocular surface squamous neoplasia (OSSN). It was a barely recognized clinical entity till almost 30 years ago, though gross conjunctival or corneal cancer was diagnosed and treated by ophthalmologists for well over a century. There was no concept of intraepithelial precursor lesions, frequency and multiplicity of presentations or biological behavior of the disease as such. Advances made in the clinical and laboratory investigations in last two decades have thrown a considerable light on this entity.
Lee and Hirst from Australia coined the term OSSN in 1995. It encompasses squamous intraepithelial neoplasia (CIN) as well as frankly invasive squamous cell carcinoma (SCC) of conjunctiva and cornea.  Both the lesions are difficult to identify clinically, hence though removed, are not always subjected to either histopathology examination or adjuvant therapy by all the ophthalmologists, therefore they are often missed for a variety of benign lesions. Recurrence rates are naturally higher in cases of OSSN (5% to 50%), which are mainly related to the adequacy of margins at initial excision.
Even though ocular surface lesions are clearly visible and easily accessible, punch biopsies from clinically evident disease are avoided to prevent scarring and loss of limbal stem cells. Repeated biopsies are certainly incompatible with its long-term good health.  Wider excision procedure of the past is often not performed today since the introduction of the topical therapy. Lack of awareness of OSSN, misinterpretation as keratoconjunctivitis or pterygium, slow growth in relatively asymptomatic patient lulls the clinician into false sense of security, with the result simple cytology investigations are not performed preoperatively. A reliable, minimally invasive and easy to perform cytology test can be of substantial value in clinical decision-making and follow-up management of OSSN. 
This technique was developed following the discovery by Egbert and co-workers in 1974, that the surface layers of conjunctival epithelium could be removed and studied by applying cellulose acetate filter paper.  Nolan, Hirst and co-workers from Australia successfully applied the technique of impression cytology for the first time in the year 1994 for diagnosis of OSSN.  Since then, many workers have used this technique the world over. Impression cytology has now been accepted as a useful noninvasive means of assessing the ocular surface epithelium and is rapidly gaining popularity in the specialized ophthalmic institutes for diagnosis of a variety of ocular surface lesions. These disorders include ocular surface squamous neoplasia, dry eye syndrome, limbal stem-cell deficiency, specific viral infections, vitamin A deficiency, allergic disorders, conjunctival melanosis, and malignant melanoma.
Until 1997 most studies used strips of cellulose acetate filter paper placed in multi-well Teflon sample holder to be transported to the laboratory. Most ophthalmologists found this procedure to be cumbersome, time consuming and therefore unsuitable. The Biopore membrane device is particularly user friendly. It offers the advantage of being quicker method, easy to apply and easy to transport with stable mechanical device.  It also ensures good adhesion of surface epithelial cells to the biopore membrane with the result; adequate specimens are obtained in a very high percentage of cases.  The sensitivity of this method is high (78-87%); the main disadvantages of impression cytology are some loss of morphological details and poor cell yield in cases of keratinizing lesions. 
In contrast to impression cytology, the scrape cytology is a traumatic procedure. Scraping the lesion collects in-situ cells. It samples much smaller area restricted to the lesion. It is more prone to the airdrying artifacts. A certain degree of expertise is required for scraping and making the smear [Table 1]. In scrape cytology smear, cell-to-cell relation is not maintained. To compensate for all these drawbacks, scrape cytology offers you a better cell yield even in keratinizing lesions and small focal lesions. It also offers good morphological details, if smear is fixed immediately. Results of scrape cytology are likely to be more specific and sensitive. Hence we prefer scrape cytology to impression cytology for diagnosis of OSSN.
Informed consent is takenPatient is made to lie in the supine position on operation table.Local anaesthetic drops are instilled in the diseased eye.Eye is kept open with self-retaining speculum. The surface of the lesion is scraped gently with the help of sterile scalpel blade (No 15) under operating microscope. Various instruments such as cyto-brush, Kimura spatula and scalpel blade can be used for scraping the lesion.Scraped material is quickly transferred to a glass slide coated with egg albumin and smeared evenly and swiftly on the slide. This step is crucial hence special attention should be given to avoid delay.The slide is immediately immersed in the jar containing 95% ethyl alcohol.The slide is stained with Papanicolaou's stain.
Application: The most common applications in diagnostic ocular pathology are:
1) Primary diagnosis of OSSN. 2) Staging and monitoring effects of treatment (topical Mitomycin C) in OSSN. , 3) Dry eye syndrome where squamous metaplasia and/or hyperkeratosis are noted. 4) Scrape cytology can also be applied for diagnosis of various conditions like specific viral infection, conjunctival melanosis and melanoma etc.
Diagnosis of OSSN
Paucity of the literature regarding criteria for diagnosis of OSSN by ocular surface cytology is highlighted by Nolan et al.  They emphasized that the ocular surface cytology does not bear strong resemblance to the cervical exfoliative cytology. However the universal cytological criteria for epithelial dysplasia can be applied to conjunctival cytology for diagnosis of OSSN.
The diagnostic criteria are:
Nuclear enlargement with raised nuclearcytoplasmic ratioHyperchromasia with coarsely clumped nuclear chromatin.Irregular nuclear membrane Nuclear pleomorphismProminent nucleoli.
Keratinizing Dysplasia/CIN: In addition to above-mentioned criteria, it shows parakeratotic cells and dyskeratotic cells with abnormal nuclei. This type of CIN is more commonly seen at ocular surface than non-keratinizing ones, which are seen commonly at other sites like cervix. Keratinisation and dyskeratosis are common features of CIN III of conjunctiva (not of CIN I / II).  Dyskeratosis is really a vital feature in ocular surface cytology, which should always alert a pathologist to look carefully for a stray dysplastic squamous cells to suspect diagnosis of keratinizing CIN. The diagnosis of non-keratinising CIN in contrast, is easy as the smear often shows sheets of dysplastic cells.
Invasive SCC: Having diagnosed dysplasia, one has to look for cytological features suggestive of stromal invasion. The distinction between CIN III and superficially invasive SCC is relatively easy in biopsy than cytology as no definite cytologic criteria reliably distinguish invasive SCC of conjunctiva from CIN III.  Though one can extrapolate criteria used in cervical cytology for diagnosis of invasive SCC, one must remember that these will be present in minority cases of conjunctival SCC, which are keratinizing and better differentiated. The criteria are cellular pleomorphism, hyperkeratinised cells, ghost cells, large number of inflammatory cells and tumour diathesis in the background. These features are less often seen in the impression cytology. Therefore a diagnosis of invasive SCC can rarely be made confidently on impression cytology, which is the serious limitation. In contrast, some of the diagnostic criteria for invasive SCC are well preserved in the scrape cytology. We could thus diagnose 9 cases of squamous carcinoma on scrape cytology out of total 11 cases (82%). In general sensitivity of scrape cytology for diagnosis of CIN is higher than that for invasive SCC. It must be noted that prominent nucleoli is not the feature to distinguish between the two as it can also be encountered in CIN.
Interpretation of Scrape Cytology Smear
Following steps, if followed systematically, would assist in arriving at the diagnosis:
Note the degree of cellularity and fixation of cells.Concentrate first on the epithelial cells.Note the patterns on low magnification and cell morphology on high magnification.Make a comment if more than one epithelial cell type is observed (goblet cells, metaplastic squamous cells, etc.).Look for stromal and inflammatory cells. Note the cell types.Now concentrate on the background. Look specifically for fluid material, pus cells, tumour diathesis, ghost cells etc.Analyze all the features and correlate with clinical impression.Finally give a composite meaningful report.Comment, when necessary, should always be added
Pitfalls in Diagnosis of OSSN
(1) Keratinising Flat Lesions: Keratinising lesions offer the biggest challenge to diagnosis in ocular surface cytology.  These include actinic keratosis, keratinizing dysplasia and a plaque like keratinising SCC. The group of leukoplakic lesions of the limbal conjunctiva, which result from prolonged exposure to UV light, is classified as actinic keratosis by W.H.O. Frequently found at the advancing edge of pterygium, these lesions appear as circumscribed grayish white transluscent area over the corneal surface. The transformed epithelium always shows acanthosis, hyperkeratosis and parakeratosis. Epithelial cells may exhibit dysplasia in some but not all the lesions of actinic keratosis. When dysplasia is recognized on cytology, that lesion should be classified as OSSN rather than actinic keratosis, which may have banal connotation to ophthalmologists.
Abundant surface keratinisation usually occurs in keratinizing type of CIN and SCC.  This may result in inadequate number of dysplastic cells in the smear, which are situated beneath the layers of keratin. The cell yield is known to be less in all hyperkeratotic lesions of conjunctiva. However it is much better in scrape cytology than impression cytology. Repeated and more vigorous scraping of the lesion can also increase cell yield. In such a situation, scrape cytology definitely scores over impression cytology. Whenever pathologist sees scanty dysplastic cells in an inflammatory background, he should correlate cytologic features with the clinical impression and if the clinical impression is 'suspicious of hyperkeratotic lesion' then the smear should be critically evaluated for parakeratotic, hyperkeratotic and dysplastic cells accompanied by inflammatory cells. When these cells, however scanty, are seen in the smear then the possibility of keratinising CIN or SCC should be suspected and excision should be recommended. Pathologist should refrain from offering a diagnosis of 'Inadequate specimen' or worse still, as inflammatory lesion in such a situation.
(2) Pseudoepitheliomatous Hyperplasia (PEH): It is a reactive proliferation of surface epithelium secondary to an inflammatory or infectious process. The epithelial cells exhibit reactive changes and increased mitotic activity mimicking CIN/SCC cytologically as well as histologically. Pathologist should distinguish reactive atypia from neoplastic atypia. Inflammatory cells and granulation tissue are commonly seen in the background. Granulomatous reaction can also be seen in case of fungal, parasitic infection or sarcoidosis and offers a definite diagnostic clue. Clinical correlation and microbiological investigations are useful steps in cases of difficulty. Rapid growth over few weeks, younger patients, lack of occupational history of exposure to UV light, reactive atypia and inflammatory cells seen on cytology aid in the distinction from OSSN. We encountered a single case of PEH diagnosed as such on excision biopsy specimen, which was labeled as benign epithelial lesion (non-dysplastic) on cytology. Awareness of this entity is necessary to avoid misinterpretation. A rule of thumb that 'Anything that grows rapidly is not OSSN and is more likely to be a benign lesion' should be applied while evaluating atypical cytologic features in ocular surface cytology.
(3) Sebaceous Carcinoma: Sebaceous carcinoma originating primarily in the eyelid or caruncle can be associated with extensive pagetoid spread within the conjunctival epithelium. If the Sebaceous Carcinoma is unknown then pagetoid spread can easily masquerade as in-Situ SCC of conjunctiva, unless the pathologist detects the vacuolated tumor cells of sebaceous carcinoma and performs fat stain on frozen section.
(4) Reactive Atypia: Cellular changes related to topical mitomycin C therapy mimic those seen following radiation therapy which may be mistaken for OSSN. These are localized in the superficial layers of the epithelium.  Though nuclei appear big and hyperchromatic, N: C ratio is not raised and the chromatin usually appears smudgy. Bi- and multinucleation is often seen. These features aid to distinguish it from recurrence of OSSN. ,
We prefer scrape cytology to impression cytology for diagnosis of OSSN. Not only the cell yield is much better in the former but also the morphological details are better preserved. The background is also better appreciated. Air-drying artifact can be avoided with experience by swiftly making smear and fixing it immediately.
34 cases of ocular surface lesions were studied by scrape cytology followed by histopathology over a period of 4 years. Out of these, cyto-histo concordance was observed in 31 cases (91%). In 3 cases, diagnosis of OSSN or malignancy was not offered on cytology. When these smears were reviewed, material was found to be inadequate in 2 out of 3 cases. These were finally diagnosed as SCC and sebaceous carcinoma each on histology. The third case showed scattered keratinized dysplastic squamous cells, which were masked by the inflammatory cells in the background and hence overlooked. Overall the scrape cytology results correlated well with histopathology and clinical impression. It is interesting to note that there was not a single false positive case in our study. The break up of scrape cytology cases (34) is given in the [Table 2].
Ocular surface lesions are easily accessible to the application of scrape cytology, which is a rapid, non-invasive, easy to perform and inexpensive technique.Keratinizing lesions offer the biggest challenge to diagnosis. Regular application of scrape cytology to ocular surface would elevate the confidence level and experience among the pathologistsDefinitive diagnosis of OSSN on scrape cytology would help ophthalmologists in clinical decision making and obviate the need for more invasive biopsy procedure for primary diagnosis and diagnosis of recurrence.
|1||Nolan GR, Hirst LW. Impression cytology in the diagnosis of ocular surface squamous neoplasia. Br J Ophthalmol 2001; 85: 888a - 888.|
|2||Nolan GR. Overview of cytology of OSSN with special reference to impression cytology, Ophthalmic pathology for general anatomic pathologists. Pathol International 2004; 54, (S1) S397-S400.|
|3||McKelvie P. Ocular surface impression cytology. Advances in anatomic pathology. 2003;10: 328-37.|
|4||Nolan GR, Hirst LW, Wright RG, et al. Application of impression cytology to the diagnosis of conjuctival neoplasms. Diagn Cytopathol 1994; 11:246-9.|
|5||Tole D, McKelvie PA, Daniell M. The reliability of impression cytology for the diagnosis of ocular surface squamous neoplasia employing the biopore membrane. Br J Ophthalmol 2001; 85: 154-8.|
|6||Salomao DR, Mathers WD, Sutphin JE, et al. Cytologic changes in the conjunctiva mimicking malignancy after topical mitomycin C chemotherapy. Ophthalmol 1999; 106:1756-61.|
|7||McKelvie PA, Mark Daniell. Impression cytology following mitomycin C therapy for ocular surface squamous neoplasia. Br J Ophthalmol 2001;85:1115-9.|
|8||Singh R, Joseph A, Umapathy T, Tint NL, Dua HS. Impression cytology of the ocular surface. Br. J. Ophthalmol 2005; 89: 1655 -9.|