Article Text

Discoid corneal oedema and high intraocular pressure following PRK
  1. Herzlia Medical Center, Herzlia, Israel
  2. Ein Tal Eye Center, Tel-Aviv, Israel
  1. Herzlia Medical Center, Herzlia, Israel
  2. Ein Tal Eye Center, Tel-Aviv, Israel
  1. Israel Kremer, MD, Ein Tal Eye Center, 17 Brandeis Street, Tel-Aviv, Israel.

Statistics from

Editor,—We present an unusual corneal oedema which was observed following scleral indentation during a buckling procedure in a patient who had undergone successful photorefractive keratectomy (PRK), and in a second patient who presented with steroid induced high intraocular pressure following PRK.


Patient 1

A 33-year-old man underwent bilateral PRK by the Visx 20/20 excimer laser, with a 6 mm ablation zone. The refraction before the procedure was −6.00/−2.00 × 177° in the right eye and −5.00/−1.75 × 180° in the left. The corrected visual acuity was 20/20. Five months after PRK the patient’s uncorrected vision was 6/9 in both eyes. With correction of plano/−0.75 × 175° in the right eye and −0.50 in the left, his vision was 20/20 in each eye. The steroid drops were stopped a month previously.

Six months after PRK, he noticed a shadow on the nasal field of his right eye and deterioration of vision. On admission the visual acuity in the right eye was 6/60 and the IOP 15 mm Hg. The cornea showed no pathology and there was no noticeable haze. The anterior chamber was deep and clear. Funduscopy revealed detachment of the temporal and inferior retina with two peripheral horseshoe tears at 9 and 6 o’clock. Examination of the left eye was unremarkable apart from a +1 paracentral arcuate corneal subepithelial haze.

During buckling procedure, at the stage of scleral indentation and localisation of the tears, while elevating the IOP close to arterial pressure, the central cornea became very hazy, to the point where it was almost impossible to visualise the retina. The corneal oedema occupied the whole 6 mm zone of PRK ablation and involved the epithelium and subepithelial layers. An attempt to scrape the oedematous epithelium over the central cornea was difficult and we realised that there was oedema deeper to the epithelium. Therefore the procedure was stopped after several strikes. Subsequently, the eye was kept soft by repeated paracentesis of the anterior chamber. The localisation of the retinal tears was continued and cryopexy performed through the hazy cornea with great difficulty. Twenty five minutes later, while placing the scleral sutures of the sponge, the cornea became clearer and the retinal tears were clearly observed. The location of the tears was double checked and found to be correct. During the rest of the surgery, the PRK zone continued to be oedematous, while the rest of the cornea was clear.

The following day, the cornea was found to be perfectly normal, except for small punctate epithelial erosions. The retina was well attached. Endothelial specular microscopy performed 1 week after the surgery revealed a normal and equal cell count of 2500 cells/mm2 in both eyes.

Patient 2

A 56-year-old man underwent PRK in his left eye for moderate myopia in July 1993. His best corrected visual acuity of that eye before PRK was 6/6(−) with correction of −5.75/−0.50 × 180°. The IOP was 18 mm Hg bilaterally; the anterior and posterior segments were unremarkable.

The patient was treated with dexamethasone (0.1%) drops six times a day following PRK, and after 6 weeks of local steroid treatment he presented with a complaint of left visual blur. On admission, his best corrected visual acuity was 6/60 in this eye. The IOP was found to be 18 mm Hg in the right eye and 52 mm Hg in the left eye. A central discoid 6 mm zone of epithelial and subepithelial oedema was observed. The anterior chamber was deep and there was no evidence of papillary block. The IOP was reduced within 2 hours to 20 mm Hg by local application of Iopidine (0.5%), Tiloptic (0.5%), and pilocarpine (2%) drops and intravenous injection of 500 mg acetazolamide. The discoid corneal oedema completely regressed within 4 hours. He was subsequently treated by Tiloptic (0.5%) drops twice daily and FML drops four times daily which were tapered off within 3 months. Final refraction, 2 years later, revealed visual acuity of 20/20(−) with −0.50 spherical correction. The cornea was found to be completely normal with no evidence of haze. Specular microscopy 2 years after PRK revealed a normal and equal cell count of 2300 cells/mm2 in both eyes.


Histopathological studies have characterised the changes occurring during healing of rabbit and monkey corneas following PRK. In these models, a subepithelial haze develops at 3 weeks after ablation and is caused by a subepithelial accumulation of active keratocytes, vacuoles containing glycosaminoglycans, and newly synthesised type III collagen fibres, with greater interfibre spacing than normal.12This subepithelial newly formed extracellular matrix2persisting for several months, may facilitate the accumulation of subepithelial fluids when IOP is significantly increased, enhancing the diffusion of water into the corneal stroma from the anterior chamber through the endothelial barrier. The newly formed hyperplastic epithelium is probably more permeable to water than its neighbouring peripheral epithelium3 and may become oedematous during this period. The absence of Bowman’s layer may also contribute to the subepithelial and intraepithelial fluid accumulation. Infrequently there is a need to remove an oedematous epithelium during buckling procedure, for better visualisation of the retina. In our two patients, it seems that both the epithelium itself and the subepithelial tissue became oedematous. The only common pathogenic mechanism among these two patients is the high IOP, which compromised the homeostasis of the PRK treated cornea. The relative rapidity of development and regression of the superficial corneal oedema may be explained mainly by the hydrostatic effect of a significantly increased IOP. It seems that with the future increase in PRK procedures in myopic patients, we will be facing a new intraoperative complication which has not yet been reported in the English literature.


View Abstract

Request Permissions

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.