Patients and methods

Our prospective study included 16 myopic patients, 10 women and six men, aged between 32 and 52 years, who presented with unilateral recalcitrant recurrent corneal erosion. Their myopia ranged between −2.0 and −7.0 dioptres and astigmatism ranged between −1.0 and −3.5 dioptres. All had recurrent corneal erosion unresponsive to our routine treatment, which included scraping off all the loose epithelium up to the limbus, patching with antibiotic ointment, and maintenance treatment by a lubricating ointment at bedtime. In 11 patients (68.7%) the corneal erosion recurred once following the first attack which was treated by scraping and a patch; in five patients (31.3%) the corneal erosion recurred twice.

Twelve patients (75%) reported a previous episode of corneal trauma in the affected eye, and in four patients (25%) the corneal erosion had occurred spontaneously.

All patients had signs of EBMD, including microcysts, intraepithelial debris, fingerprint patterns or map patterns, at least in one eye. The duration of symptoms of the epithelial basement membrane disease, such as pain upon waking, visual blur, and photophobia, ranged between 9 and 26 months with a mean of 18 months. The follow up time after excimer laser treatment ranged from 26 to 42 months with a mean of 31 months (Table 1).

Each patient underwent both subjective and cycloplegic refraction, achieving the best corrected visual acuity. However, when significant corneal surface irregularity was present, the cylinder and spherical error to be treated was calculated according to the patient’s present spectacle data.

The method of combined PRK and PTK performed was as follows: the whole epithelium was scraped up to the limbal zone with a Beaver blade; PRK was subsequently performed with the Visx 20/20 excimer laser at a 6 mm diameter ablation zone centred on the pupil, for correction of the spherical component of the myopia. We attempted to correct both the myopia and astigmatism according to the preoperative refraction. When astigmatism was also present, it was corrected with a 4.8 mm slit width for cylinders up to −2.5 dioptres, and with a 5.0 mm slit width for cylinders between −2.5 and −3.5 dioptres. Subsequently, the laser computer program was changed to the PTK mode and the treatment zone diameter was programmed at 3.0 mm with a depth of 6 μm. Thereafter, the whole area of Bowman’s layer peripheral to the PRK zone, was treated up to the limbus, by confluent 3.0 mm diameter circles of PTK at a depth of 6 μm (Fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

PRK-PTK treatment scheme illustrating the central 6 mm diameter zone of PRK, surrounded by confluent 3 mm diameter circular zones of PTK, reaching the corneal limbus (L).

Following treatment, the eye was patched with chloramphenicol (5%) ointment. On the following day, garamycin, Voltaren (diclofenac), and dexamethasone drops were applied together with chloramphenicol ointment, three times a day, until complete epithelialisation. Thereafter, dexamethasone drops were administered five times a day for 1 month, and then gradually tapered off within 5 months.

Results

The duration of re-epithelialisation ranged between 7 and 9 days with a mean of 8.5 days. All patients showed complete alleviation of their symptoms 1 month after treatment, with no recurrence of corneal erosion during their follow up. The uncorrected visual acuity following treatment ranged between 6/9 and 6/12. The postoperative best corrected visual acuity (BVA) did not change compared with the preoperative BVA in six eyes, improved by at least two Snellen lines in eight eyes, and was reduced by one Snellen line in two eyes (Table 1). It should be noted that in those eight patients in whom BVA improved after treatment, the loose epithelium and/or anterior basement membrane pathology were located centrally. Therefore, the preoperative subjective refraction was inaccurate, and PRK had to be performed according to the previous spectacle prescription. The final refraction following treatment ranged between +0.25 and −1.0 dioptres spherical equivalent (Table 1). Visual rehabilitation was gradual and relatively slow, as the post-treatment BVA reached its maximal level within a period of 5–6 months (mean 5.2 months). All patients underwent PRK in the other eye within a period of 4–6 months following treatment of the affected eye. A +1 diffuse faint subepithelial haze was still present at the end of the follow up period in those patients with a spherical equivalent higher than −4.12 dioptres (nine out of 16 eyes).

Discussion

There are several reports on the treatment of recalcitrant recurrent corneal erosion syndrome by excimer laser PTK.23-31 The reported rate of success, regarding alleviation of symptoms and prevention of recurrence of epithelial erosion, ranges between 74.4% and 100%.23-31 Dauschet al 23 reported a success rate of 74.4% in 74 eyes treated by PTK. O’Brart et al 27reported that four out of 17 eyes (23.5%) experienced recurrent episodes 3–6 months after PTK and two have subsequently been retreated, one of whom has been symptom free for 12 months following retreatment. Postoperative BVA was unaltered in nine eyes and improved by at least one Snellen line in eight eyes. Both Forster et al 31 and Algawi et al 24 reported that none of their treated patients had experienced recurrence of corneal erosion. However, Algawi et al24 reported that two of 14 patients still complained of transient, recurrent foreign body sensation following PTK, although none of their patients developed recurrence of the erosion.

Öhman et al 26 reported that 20 of 76 treated eyes (26.3%) had recurrences following PTK. Thirty five of their patients had no symptoms after treatment and 21 patients had minor symptoms. In all the above mentioned studies,23-31the method of PTK included focal ablation restricted to the affected zone of corneal epithelium, at a maximal depth of 40 μm whenever there was no actual erosion. Alternatively, when a true erosion was present together with loose epithelial margins, the corneal area not covered with epithelium was treated superficially at an ablation depth between 1 and 5 μm, in order to interfere as little as possible with Bowman’s layer.23 Subsequently, the loose epithelium beyond the margins of the erosion was treated directly within an area of one laser spot size of 1.5 mm, at a depth of 30–40 μm.23 Algawi et al 24 and O’Brart et al 27 reported that they mechanically removed the locally loose epithelium, and subsequently a 20–30 pulse ablation was applied to Bowman’s layer, achieving an ablation depth of 5–7 μm. Recently, Kaplan-Messas et al 32 reported, for the first time, a study comparing the results of the above described focal mode of PTK treatment with the results of a diffuse mode of PTK, whereby they shot 54 pulses spread over the entire Bowman’s layer following total epithelial debridement, in nine overlapping 6 mm diameter spots with a mean ablation depth of 2 μm. Patients were monitored for refractive and topographic changes. Two eyes with focal PTK, but no eye with diffuse PTK had recurrent ulcers after a mean follow up of 8.0 (SD 2.49) and 9.03 (2.62) months, respectively. A final hyperopic shift was found, but it was not significantly different between the two groups. The vectorial change of astigmatism was significantly lower in the focal PTK group, but surface regulatory index changes were not significantly different between the two groups.32

Interestingly, the diffuse PTK method utilised by the latter authors32 is quite similar to our method of combined PRK and PTK, which actually includes pancorneal ablation. However, our method is different in the sense that the central 6 mm ablation depth is much deeper as PRK, and not PTK, is performed in this zone. The stromal ablation depth in the central 6 mm zone of PRK depends on the degree of myopia.

Lind et al,33 in a recent presentation on excimer laser treatment for RCE, reported that the overall refractive change in 43 PTK treated patients was a hyperopic shift of 0.69 dioptres. These authors33 did not mention whether they utilised the diffuse or focal mode of PTK. Kaplan-Messas et al 32 found a hyperopic shift of 0.87 (0.81) dioptres in the diffuse treatment group and 0.38 (0.74) dioptres in the focal treatment group. However, there was no statistically significant difference between the two groups. In our group of patients, there was only one eye out of 16 (6.2%) which had a post-treatment spherical equivalent of +0.25 dioptres; the remainder (15 eyes) had a spherical equivalent ranging between plano and −1.0 dioptre. Fourteen out of 16 eyes (87.5%) had a final myopic spherical equivalent less or equal to −1.0 dioptre. This small myopic shift compared with the hyperopic shift found in previous studies3233 may be related to the peripheral confluent PTK treatment, surrounding the central zone of PRK or to the inaccurate pre-PRK refraction performed in those patients with central RCE. As PRK was not performed in these two studies3233 it is impossible to compare their refraction results and our refraction results following treatment. Additionally, the number of patients included in our study is relatively small compared with the latter two studies.3233

With regard to our excimer laser ablation technique, it should be noted that Gibralter and Trokel34 reported their excimer ablation technique, developed in order to create a regular optical surface. They customised an ablation programme using circular PTK elements of varying diameters, to flatten identified high areas on corneal topography, and a standard myopic ablation to achieve central flattening and elimination of residual myopia. These two techniques are quite similar, although they were utilised for the different corneal disorders.

Pertaining to the high rate of success of our above described method of treatment, we postulate that the diffuse pancorneal ablation induces a better adhesion capacity of the whole basal epithelium to the stromal bed. It has been shown that following excimer laser ablation, a normal structured basement membrane with its anchoring filaments and adhesion proteins finally develops.35 We believe that recalcitrant RCE is a diffuse disease of the corneal epithelium and its basement membrane, although it quite often presents as a local abnormality. Therefore, it should be treated extensively either by the diffuse PTK method described by Kaplan-Messas et al,32 or by the combined PRK and PTK method which was performed in our above presented myopic patients. Unfortunately, we did not have a control group of RCE patients treated by focal PTK, with whom we could compare our results.

It is important to note that Fountain et al 35 have found that following excimer laser keratectomy, the anchoring fibrils, hemidesmosomes, and the basal lamina do not completely normalise even after 15 months.35Therefore, we are of the opinion that it is imperative to follow these patients for at least 15 months after diffuse PTK, before drawing any conclusions regarding the effect of this treatment on the recurrence of corneal epithelial erosion.