Specular Microscopy

Specular microscopy takes magnified images of the corneal endothelium, the single layer of cells on the inner surface of the cornea. These cells pump fluid out of the cornea to keep it clear and thin. They do not divide in adults, so when cells are lost, the neighbouring cells enlarge and spread to cover the gap. If the cell count falls too low, the pump fails and the cornea swells and loses its clarity.

The instrument shines light on the cornea and records the light that is reflected from the boundary between the endothelium and the aqueous humour (specular reflection). The cell borders appear dark and the cell bodies bright, in a pattern like a honeycomb.

Non-contact instruments are used mostly, with automated analysis of a number of cells, and contact and wide field instruments are used in eye banks and research. The software outlines the cells and reports the endothelial cell density (cells per mm²), the coefficient of variation of cell area (polymegathism), the percentage of hexagonal cells (pleomorphism), and the central corneal thickness.

The test takes 2 to 5 minutes. Cell density is about 3000 to 4000 cells per mm² in young adults, about 2500 in middle age, and about 2000 to 2500 in older people, with a loss of about 0.3 to 0.6% a year. The cornea usually stays clear above 1000 cells per mm², and oedema is likely at lower than 500 cells per mm². Cataract surgery causes a loss of about 5 to 10% of cells.

Indications.

• Before intraocular surgery (cataract surgery, glaucoma surgery, phakic lens implants), particularly in eyes with risk factors, to assess the risk of corneal decompensation.
• Corneal oedema and unexplained loss of clarity after surgery, and assessment of bullous keratopathy.
• Fuchs endothelial dystrophy and other endothelial disease (posterior polymorphous dystrophy, iridocorneal endothelial syndrome), for diagnosis and monitoring.
• Assessment of donor corneas in eye banks, and monitoring of corneal grafts (rejection and graft failure).
• Contact lens wearers with long term wear, and effects of hypoxia (polymegathism).
• After trauma, uveitis, high intraocular pressure, or angle closure, which can damage the endothelium.
• Monitoring of drug effects, and of glaucoma drainage tubes close to the cornea.
• Assessment of the cornea before laser refractive surgery in suspect eyes.

Not possible when the cornea is too cloudy or scarred to image, or when the view is blocked by oedema, though in those cases confocal microscopy may help.

Contraindications and precautions.

• No contraindications for non-contact methods.
• Contact methods are avoided in infection, open globe injury, or fragile corneas, and need anaesthetic drops.
• Poor results in patients with poor fixation, nystagmus, severe dry eye, a small palpebral aperture, or marked corneal opacity.
• Corneal oedema and scars hide the cell borders, and no valid count is obtained.
• Contact lenses are removed before the test, as lens wear can change the endothelium image for a short time.
• Tonometry and anaesthetic drops are given after the test, since they affect the surface and image.

Preparation and technique.

The patient sits at the instrument with the chin and forehead supported. The operator aligns the camera on the central cornea and asks the patient to look at a target and keep the eye open. The instrument focuses automatically on the endothelium and captures an image.

The operator selects the cells to analyse (at least 75 to 100 cells for a reliable result, and the centre of the cornea is tested), and the software gives the cell density and shape indices. Several images are taken in each eye, and the best image or the average is used. Peripheral and paracentral images may be added in graft or localised disease.

The result is checked by looking at the image for correct cell outlines.

Normal appearances.

The image shows a regular honeycomb of hexagonal cells of similar size, with dark, thin borders. Cell density is about 3000 to 4000 cells per mm² in young adults, with a coefficient of variation of cell area of 0.30 or less, and 60 to 70% hexagonal cells or more. Mean cell area is about 250 to 350 microm². Corneal thickness is about 520 to 560 micrometres. In children, density is over 3500 to 4000 cells per mm².

Interpretation.

• Low endothelial cell density: under about 2000 cells per mm² suggests reduced reserve, and surgery needs care. Under about 1000 cells per mm² carries a high risk of corneal oedema after surgery, and under 500 cells per mm² usually shows decompensation.
• Polymegathism (increase in coefficient of variation to 0.40 or more) and pleomorphism (hexagonal cells under 50%) show endothelial stress, such as contact lens wear, diabetes, after surgery and with ageing.
• Fuchs endothelial dystrophy: dark round structures (guttae) that distort the mosaic, loss of cell borders, enlarged cells, and low density. Guttae appear as dark spots with bright centres.
• Iridocorneal endothelial syndrome: dark cells with a reversed light and dark pattern ("ICE cells"), and an abnormal, often one sided, mosaic.
• Posterior polymorphous dystrophy: vesicles, bands and abnormal cells.
• After surgery or trauma: cell loss with enlarged, irregular cells, and a slow recovery by cell spread. Graft failure: falling density, and a count of under about 500 to 700 cells per mm² predicts failure.
• Corneal oedema: dark, blurred image with lost borders, so counts cannot be measured.
• Inflammatory cell deposits (keratic precipitates) are seen as bright or dark round structures in uveitis.
• Results are read together with the corneal thickness and with the slit lamp findings.

Critical values and findings needing immediate action.

• There are no emergency values.
• Very low endothelial cell density (under about 1000 cells per mm²) before cataract or other intraocular surgery: referral to a corneal specialist, adjustment of surgical technique, or combined corneal and cataract surgery.
• Falling cell counts in a graft, with new corneal swelling or signs of rejection (redness, pain, reduced vision): urgent review for rejection, in the first days.
• Acute corneal oedema with pain and reduced vision: urgent assessment for high pressure, infection and endothelial failure.
• Donor tissue under the acceptable cell count in an eye bank: not used for transplantation.

Limitations and artefacts.

• Image quality depends on corneal clarity, tear film, and fixation. Corneal oedema, scars, deposits, and haze reduce or prevent counting.
• Only a small central area (a fraction of a square millimetre) is sampled, and cells in the periphery may differ, so the count may not represent the whole cornea, particularly in localised disease.
• Automated software can outline cells wrongly, and errors in cell borders change the count, so manual correction and checking of the image matter.
• Different instruments and methods give different results, so serial measurements use the same device.
• Cell density alone does not give function. Corneas with low counts may stay clear, and corneas with higher counts and poor pump function may swell. Thickness and clinical findings complete the picture.
• Cell loss with ageing and with contact lens wear is gradual, and a single measurement does not give the rate.
• Normal values vary with age, and the reference range is set by age.

Special groups.

• Older patients before cataract surgery: baseline counts identify those at risk of corneal failure.
• Patients with Fuchs dystrophy: count, thickness and guttae density help to decide the timing of surgery, and to select endothelial transplantation or combined surgery.
• Corneal graft recipients: counts are followed for years, and a fall signals rejection or chronic failure.
• Patients with glaucoma tubes or phakic implants: regular counts to identify progressive loss.
• Long term contact lens wearers: assessment for endothelial stress when symptoms occur.
• Donor assessment: eye banks use counts as one criterion for use of tissue.
• Regions: specular microscopes are available at eye banks and a few corneal centres. Corneal blindness from infection, trauma, vitamin A deficiency, and other causes is a leading cause of blindness, and cataract surgery by small incision techniques in many places is done without endothelial counts, with attention to surgical technique and the use of protective agents. Slit lamp examination with attention to the guttae and corneal thickness gives an approximate assessment where there is no specular microscope.

Follow up tests.

• Corneal pachymetry for thickness.
• Slit lamp examination for guttae, oedema, and keratic precipitates.
• Anterior segment optical coherence tomography.
• Confocal microscopy where specular views are not possible.
• Intraocular pressure measurement and gonioscopy.
• Repeat specular microscopy at 6 to 12 month intervals, or after surgery.
• Referral for endothelial keratoplasty when counts and function fall.