Impression cytology is a minimally invasive test that applies cellulose acetate filter paper to the ocular surface to collect superficial epithelial cells for histological, immunohistochemical, and molecular biological analysis.
In 1977, Egbert et al. established the technique for studying conjunctival goblet cells, and later Tseng improved it, leading to widespread clinical application for ocular surface diseases.
Pathological diagnosis is broadly divided into histology and cytology. Cytology is considered an adjunct to histology because it does not provide information on tissue structure, but in the eye, cytology is highly useful because histology is more invasive.
Impression cytology uses filter paper to peel off the epithelium and collect cells as a two-dimensional sheet. Conjunctival smears are collected by scraping with a cotton swab or cytology spatula. Impression cytology is superior for evaluating the distribution and density of goblet cells.
Since impression cytology is a testing method rather than a disease that examines “symptoms,” this section describes the findings observed in collected specimens and their clinical significance.
Impression cytology is applied to a wide range of diseases and conditions as follows.
Ocular surface diseases
Tumors
Ocular surface squamous neoplasia is classified into conjunctival intraepithelial neoplasia (CIN) and invasive squamous cell carcinoma. The main risk factors are as follows.
Compare the types of filter paper.
Cellulose acetate membrane
Millipore filter with 0.22 μm pore size: The standard choice considered best.
Shape: Cut into 5 mm strips (one end square, one end tapered). Mark the surface for orientation.
For PCR (polymerase chain reaction): Use a circular filter with a diameter of 10 mm.
Nitrocellulose membrane
Trapezoidal strip (approximately 3×6 mm): Standard shape.
Need to distinguish front and back: The smooth side is the front. Adhere the back (rough side) to the conjunctival epithelium.
Immunohistochemistry: Nitrocellulose membrane is suitable.
Biopore membrane
Application for diagnosis of limbal stem cell deficiency: It is an option along with cellulose acetate and nitrocellulose.
Common characteristics: Can collect 1 to 3 layers of surface epithelium.
The collection site becomes an epithelial defect, which may cause eye pain after returning home. Explain this to the patient beforehand and prescribe corneal protective eye drops.
Because an epithelial defect occurs at the collection site, patients may experience eye pain after returning home. Explain this before the examination and prescribe corneal protective eye drops to manage it. In most cases, the defect heals naturally within a few days.
The impression cytology findings for each disease are shown below.
Some consider impression cytology to be the gold standard for diagnosing limbal stem cell deficiency (although it is not widely used in clinical practice).
The appearance of goblet cells on the cornea is direct evidence of conjunctivalization and provides a basis for diagnosis. Apply a filter to the central cornea to check for the presence of goblet cells.
The predictive accuracy of impression cytology for each tumor diagnosis is shown below.
| Disease | Diagnostic Predictive Accuracy |
|---|---|
| Ocular surface squamous neoplasia (histological diagnosis) | 80% |
| Moderate dysplasia | 77% |
| Acanthamoeba keratitis | 94.6% |
The gold standard for definitive diagnosis of ocular surface squamous neoplasia is excisional biopsy. Impression cytology and exfoliative cytology cannot reliably distinguish between superficial and invasive lesions and are not used for definitive diagnosis.
In ocular surface squamous neoplasia arising from chronic vernal keratoconjunctivitis, differentiation from pseudoepitheliomatous hyperplasia is difficult. If pleomorphic pigmented epithelial cells are detected on impression cytology, it may raise suspicion for ocular surface squamous neoplasia 1).
The accuracy of predicting histological diagnosis is 80%, and excisional biopsy is required for definitive diagnosis. Because impression cytology cannot reliably distinguish between invasive and intraepithelial lesions, it is considered an adjunctive diagnostic tool.
The cellulose acetate filter works by peeling off the outermost layers (1–3 layers) of the ocular surface epithelium along with mucous secretions.
When the filter is firmly adhered, it can collect several layers of stratified cell sheets, allowing two-dimensional evaluation of cell morphology, distribution, and density.
In normal conjunctiva, goblet cells are distributed throughout the conjunctiva. Chronic inflammation and severe dry eye destroy goblet cells and cause epithelial metaplasia (squamous metaplasia). In limbal stem cell deficiency, depletion of limbal stem cells leads to loss of the ability to replenish corneal epithelium, resulting in “conjunctivalization” where conjunctival epithelium covers the corneal surface. Identification of goblet cells on the cornea by impression cytology is evidence of this conjunctivalization.
Although impression cytology is minimally invasive, it causes epithelial defects at the collection site and is therefore not a completely noninvasive test. In recent years, the following new modalities have been developed.
The characteristics of each diagnostic modality are shown below.
| Modality | Characteristics | Remarks |
|---|---|---|
| Impression cytology | Allows evaluation of goblet cells and immunohistochemistry | Evaluates only 1–3 layers; epithelial defects present |
| In vivo confocal microscopy | Quantifies cells with 1–10 μm resolution | Requires patient cooperation for 5–15 minutes |
| High-resolution optical coherence tomography | Distinguishes invasive from intraepithelial lesions | Epithelial thickness >120 μm suggests invasion |
In vivo confocal microscopy allows observation of corneal and conjunctival cells at a resolution of 1 to 10 μm in living tissue. It enables quantification of cells but requires patient cooperation for 5 to 15 minutes.
High-resolution optical coherence tomography (high-resolution anterior segment OCT) is considered useful for distinguishing between invasive and intraepithelial lesions by identifying epithelial thickening (greater than 120 μm).
For ocular surface squamous neoplasia, several pharmacotherapies including mitomycin C 0.04%, 5-fluorouracil, interferon alpha-2b, anti-VEGF agents, and cidofovir have been investigated, but official guidelines have not yet been established.
In a 24-year-old male with ocular surface squamous neoplasia arising from chronic vernal keratoconjunctivitis, impression cytology detected pleomorphic pigment-containing epithelial cells, and tumor regression was confirmed after two cycles of mitomycin C 0.04% therapy. Subsequent excisional biopsy (including 4 mm margins, no-touch technique, cryotherapy, and amniotic membrane transplantation) revealed mild to moderate dysplasia, and no recurrence was observed during 6 months of follow-up 1). Impression cytology is also considered applicable for monitoring after mitomycin C treatment.
In vivo confocal microscopy and high-resolution optical coherence tomography are evolving non-invasive modalities. High-resolution OCT is considered useful for distinguishing invasive lesions based on epithelial thickening (>120 μm), but both still involve research-stage elements. Impression cytology retains its unique value in its ability to identify goblet cells and perform immunohistochemical staining.
