Retreatment after laser in situ keratomileusis1

Retreatment after Laser In Situ Keratomileusis Juan J. Pe´rez–Santonja, MD, Marı´a J. Ayala, MD, Hani F. Sakla, MD, Jose´ M. Ruı´z–Moreno, MD, Jorge L. Alio´, MD Objective: To evaluate the effectiveness, predictability, and safety of laser in situ keratomileusis (LASIK) retreatment for correcting residual myopia. Design: Retrospective noncomparative case series. Participants and Intervention: Fifty-nine consecutive eyes (43 patients) underwent LASIK retreatment at 3 or 6 months after the primary LASIK procedure. Lifting the corneal flap and reablating the stromal bed with a VISX 20/20 excimer laser was the procedure used for LASIK enhancement. Main Outcome Measures: The following parameters were studied before and after retreatment: visual acuity, refraction, videokeratography, applanation tonometry, and corneal thickness. Complications after LASIK enhancement also were evaluated. Follow-up was 12 months. Results: Before retreatment, only 3.38% of eyes (2 of 59) had an uncorrected visual acuity of 0.5 (20/40) or better, and after retreatment, this percentage increased to 60% (30 of 50) at 6 months and 61.8% (34 of 55) at 12 months. After reoperation, mean best-corrected visual acuity improved by half a line over the values before retreatment. The preretreatment refraction of ⫺2.92 ⫾ 1.22 diopters (D) (mean ⫾ standard deviation) decreased significantly to ⫺0.44 ⫾ 0.80 D at 6 months and to ⫺0.61 ⫾ 0.82 D at 12 months (P ⬍ 0.001). In 82% of eyes (41 of 50) at 6 months and 81.8% (45 of 55) at 12 months, the spherical equivalent was within 1.00 D of emmetropia. There was a significant regression of effect (0.38 D) between 3 and 12 months (P ⬍ 0.01). Postretreatment refraction was related to the original refraction before the primary LASIK, the preretreatment refraction, and the ablation diameter used. Although no vision-threatening complications were found, epithelial ingrowth and flap melting were more common after than before LASIK retreatment, with 31% of eyes at 12 months with epithelial ingrowth and 10.9% with flap melting. However, LASIK enhancement improved decentration and night-vision problems. Conclusions: LASIK retreatment was an effective and predictable procedure for correcting residual myopia. Epithelial ingrowth and flap melting were more frequent after than before LASIK retreatment, whereas decentration and night-vision symptoms improved. Ophthalmology 1999;106:21–28 Laser in situ keratomileusis (LASIK) is a new technique for the correction of moderate-to-high myopia.1– 4 This procedure combines lifting a corneal flap with a microkeratome and refractive photoablation on the stromal bed by means of a 193-nm argon fluoride excimer laser.1,2 Recent clinical studies found that LASIK offers good results in cases of moderate and high myopia, although clinical data on this procedure are still limited.1,3– 6 Scant information is also available concerning complications after LASIK. Although several complications have been reported,1–7 undercorrection is the most common one after this procedure. Undercorrections greater than 1.00

diopter (D) have been reported in 11.3% to 59% of cases, depending on preoperative myopia.1– 6,8 Radial keratotomy and lifting the corneal flap with another laser ablation have been suggested as procedures for the management of LASIK undercorrections9,10; however, limited data are available on these techniques.10 The purpose of the current study was to evaluate the effectiveness, predictability, and safety of LASIK retreatment for correcting residual myopia in a series of 59 eyes, with a follow-up of 12 months.

Materials and Methods Originally received: September 8, 1997. Revision accepted: July 8, 1998. Manuscript no. 97585. From the Refractive Surgery and Cornea Unit, Alicante Institute of Ophthalmology, University of Alicante School of Medicine, Alicante, Spain. Presented in part at the American Academy of Ophthalmology annual meeting, San Francisco, October 1997. The authors have no proprietary interest in any of the materials described in this article. Reprint requests to Juan J. Pe´rez–Santonja, MD, Alicante Institute of Ophthalmology, Avd de Denia 111, 03015-Alicante, Spain.

From September 1994 through December 1995, 393 LASIK procedures for correcting myopia from ⫺8.00 to ⫺22.00 D were performed by the same surgeon (JJP-S) at Alicante Institute of Ophthalmology, Alicante, Spain. The results of the first 143 consecutive eyes from this series were reported elsewhere.6 Sixty-nine eyes (17.5%) of 393 were undercorrected by 1.00 D or more, and 59 eyes (15%) underwent LASIK retreatment for correcting this residual myopia. In ten undercorrected eyes (2.5%), the patients were satisfied with the results and decided not to undergo further treatment.

21

Ophthalmology Volume 106, Number 1, January 1999 The LASIK retreatment group comprised 59 consecutive eyes (43 patients) that underwent LASIK enhancement by the same surgeon (JJP-S) at 3 (n ⫽ 16) or 6 (n ⫽ 43) months after the primary LASIK procedure. Twenty-four eyes (40.6%) were in men and 35 eyes (59.3%) were in women. Mean patient age was 32.1 ⫾ 8.5 years (mean ⫾ standard deviation [SD]; range, 21–52 years). Mean preoperative refraction before the primary LASIK was ⫺13.39 ⫾ 2.8 D (range, ⫺8.00 to ⫺19.75 D), with 19 eyes between ⫺8.00 and ⫺11.99 D (group A), 28 eyes between ⫺12.00 and ⫺15.99 D (group B), and 12 eyes between ⫺16.00 and ⫺20.00 D (group C). Mean preoperative refraction before LASIK retreatment was ⫺2.92 ⫾ 1.2 D (range, ⫺1.00 to ⫺6.75 D). Patient selection criteria for LASIK retreatment were residual myopia equal or greater than ⫺1.00 D at 3 or 6 months after primary LASIK procedure, normal anterior segment, and normal peripheral retina or treatment with photocoagulation when necessary. In addition, the cornea had to be thick enough so that the remaining total corneal thickness after retreatment was greater than 370 ␮m (⬎200 ␮m of stroma remaining under the flap).6 Informed consent was obtained from all patients after they received a detailed description of LASIK retreatment and a thorough review of LASIK-known risks. The primary LASIK procedure criteria were reported elsewhere.6 Before and after retreatment, patients underwent a complete ophthalmic examination. This examination included visual acuity, manifest and cyclopegic refractions, videokeratography (EyeSys Corneal Analysis System, Houston, TX), slit-lamp microscopy, Goldmann applanation tonometry, indirect ophthalmoscopy, and corneal thickness (DGH-500 pachymeter; DGH Technology, Inc, Exton, PA). Postoperative examinations were conducted at 3, 6, and 12 months. Five of the 59 eyes in the study were unavailable for the 3-month follow-up; 9 eyes, for the 6-month follow-up; and 4 eyes, for the 12-month follow-up. All primary LASIK procedures were performed with the patient under topical anesthesia using the Automated Corneal Shaper microkeratome (Chiron Vision, Irvine, CA) and the VISX 20/20 193-nm, argon–fluoride excimer laser (VISX Inc, Santa Clara, CA).6 A nasally based corneal flap, 130- (n ⫽ 23) or 160-␮m (n ⫽ 36) thick and 8.5 mm in diameter, was performed using the automated microkeratome. The exposed stromal bed was then ablated by using the VISX 20/20 laser, with an energy fluence of 160 mJ/cm2 and a frequency of 6 Hz. A multizone approach was used, with 50% of correction done at the first zone, 30% at the second, and 20% at the third. The triple ablation profile was 4 mm (first zone)/4.5 mm (second)/5 mm (third) (n ⫽ 7), 4.5/5.0/5.5 mm (n ⫽ 13), or 5.0/5.5/6.0 mm (n ⫽ 39), depending on the intended correction and the preoperative corneal thickness, so that the resulting central corneal thickness after surgery was greater than 370 ␮m. The mean calculated central ablation depth was 129 ⫾ 21 ␮m (range, 84 –178 ␮m). After ablation, the flap was replaced without sutures. Eyes were not occluded after the procedure. Antibiotic (tobramycin 0.3%, Tobrex; Alcon S.A., Madrid, Spain) and corticosteroid (fluorometholone 0.1%, FML; Allergan S.A., Madrid, Spain) eyedrops were instilled four times a day for the first 10 days. LASIK retreatments were performed by lifting the flap and reablating the stromal bed. Before surgery, the patient was taken to the slit lamp and the edge of the flap was marked with gentian violet on the temporal side. The procedure was performed with the patient under topical anesthesia with 0.4% oxibuprocaine. The cornea was marked with two additional pararadial gentian violet lines, and a small area of epithelium on the temporal side was debrided using a blunt spatula. A flat spatula was then inserted beneath the corneal flap edge. Once the flap interface was identified, the spatula was passed along the flap edge circumferentially, and the interface was dissected. The hinged flap was lifted using a

22

nontoothed forceps and placed against the nasal sclera. The stromal bed was then ablated using the VISX 20/20 excimer laser. There was no specific retreatment nomogram. A single-zone ablation was used, which was 5.0 mm (n ⫽ 12), 5.5 mm (n ⫽ 12), or 6.0 mm (n ⫽ 35) in diameter, depending on the intended correction and the preoperative corneal thickness. The mean calculated central ablation depth was 37 ⫾ 13 ␮m (range, 18 –94 ␮m). After ablation, the flap was replaced to its original position, and the interface was irrigated copiously with balanced salt solution using a 23G cannula for removing debris and residual epithelial cells. The flap was then centered for proper alignment according to the pararadial marks, and the keratectomy incision was dried with absorptive sponges (Surgical Spears; Merocel Corp., Mystic, CT). After 3 minutes, the flap was checked for adhesion by depressing the peripheral host cornea and watching to ensure that the resulting indentation went into the flap. When adhesion was confirmed, the procedure was finished. Eyes were not occluded after retreatment, and the postoperative regimen was identical to that after the primary LASIK procedure. Both parametric and nonparametric statistical analyses were done based on the distribution of the data under consideration. Group differences for continuous variables were tested using the unpaired Student’s t test, one-way analysis of variance (one-way ANOVA), or two-way analysis of variance (two-way ANOVA) for normally distributed data, and the Wilcoxon signed rank, Mann– Whitney, and Kruskall–Wallis tests for non-normally distributed data. Differences for categoric variables were tested using the chi-square or Fisher exact test for independence. Differences for ordered categoric variables were tested using the Wilcoxon and Mann–Whitney tests. Paired proportions were compared using the McNemar test. Correlations between continuous variables were obtained using Pearson’s correlation coefficient. Differences were considered statistically significant when P values were less than 0.05.

Results Visual Acuity To evaluate visual acuity, we must take into account the reduced levels of uncorrected- and best-corrected visual acuity (BCVA) in patients with high myopia due to myopic chorioretinal degeneration. Mean uncorrected visual acuity (UCVA) before the initial treatment was 0.022 ⫾ 0.012 (mean ⫾ standard deviation), and at the time of retreatment, it was 0.19 ⫾ 0.10. After retreatment, mean UCVA increased to 0.51 ⫾ 0.16 at 3 months, 0.50 ⫾ 0.17 at 6 months, and 0.50 ⫾ 0.17 at 12 months (Fig 1). UCVA after retreatment was significantly better than the values before retreatment at all follow-ups (two-way ANOVA, P ⬍ 0.001). There were no significant differences between mean visual acuities at the three follow-up times (two-way ANOVA, P ⫽ 0.297). At the time of retreatment, 3.38% of eyes (2 of 59) had a UCVA of 0.5 (20/40) or better, and after retreatment, 59.2% of eyes (32 of 54) at 3 months, 60% (30 of 50) at 6 months, and 61.8% (34 of 55) at 12 months had a UCVA of 0.5 or better. Mean BCVA before the primary LASIK procedure was 0.51 ⫾ 0.16, and at the time of retreatment, it was 0.54 ⫾ 0.15. After retreatment, mean BCVA increased to 0.60 ⫾ 0.14 at 3 months, 0.59 ⫾ 0.15 at 6 months, and 0.59 ⫾ 0.16 at 12 months (Fig 1). It improved by 0.06 at 3 and 6 months and by 0.054 at 12 months (0.1 ⫽ one line) over the values before retreatment. BCVA after retreatment was significantly better than the values before retreatment at all follow-ups (two-way ANOVA, P ⬍ 0.001). There were

Pe´rez–Santonja et al 䡠 Retreatment after LASIK

Figure 1. Time course of uncorrected visual acuity and best-corrected visual acuity (mean ⫾ standard deviation). Numbers in parentheses indicate number of eyes studied at each period.

no significant differences between mean BCVAs at the three follow-up times (two-way ANOVA, P ⫽ 0.401). No patient lost two or more lines of BCVA. Our data show an improvement in both UCVA and BCVA after LASIK retreatment and stability in both visual acuities after the third month.

Refraction Mean spherical equivalent refraction before the initial treatment was ⫺13.39 ⫾ 2.81 D (range, ⫺8.00 to ⫺19.75 D), and before retreatment, it was ⫺2.92 ⫾ 1.22 D (range, ⫺1.00 to ⫺6.75 D). After retreatment, mean refraction was ⫺0.18 ⫾ 0.80 D (range, ⫹2.50 to ⫺2.25 D) at 3 months, ⫺0.44 ⫾ 0.80 D (range, ⫹1.25 to ⫺3.00 D) at 6 months, and ⫺0.61 ⫾ 0.82 D (range, ⫹1.00 to ⫺3.50 D) at 12 months (Fig 2). The differences between values before and after retreatment were statistically significant at all follow-ups (two-way ANOVA, P ⬍ 0.001). There was a significant regression of effect (0.21 D) between 3 and 6 months after retreatment (two-way ANOVA, P ⬍ 0.01) and also between 6 and 12 months (0.17 D) (P ⬍ 0.01). The whole regression between 3 and 12 months was 0.38 D (P ⬍ 0.01) (Fig 3). In 92.6% of eyes (50 of 54) at 3 months, 82% (41 of 50) at 6 months, and 81.8% (45

Figure 3. Time course of refractive outcome (spherical equivalent) (mean ⫾ standard deviation).

of 55) at 12 months, the spherical equivalent after retreatment was within 1.00 D of emmetropia. Mean keratometry and pachymetry at the time of retreatment were not correlated with postoperative refraction (keratometry: r ⫽ 0.05, P ⫽ 0.768 at 6 months; r ⫽ 0.01, P ⫽ 0.962 at 12 months. Pachymetry: r ⫽ ⫺0.14, P ⫽ 0.341 at 6 months; r ⫽ ⫺0.19, P ⫽ 0.159 at 12 months). Moreover, no differences were found in postoperative spherical equivalent between eyes operated on at 3 months after the primary LASIK procedure and those operated on at 6 months (Student’s t test for unpaired data, P ⫽ 0.632 at 3 months; P ⫽ 0.714 at 6 months; P ⫽ 0.665 at 12 months). However, at 12 months, those eyes from the initial treatment group C (mean refraction, ⫺1.21 ⫾ 1.16 D) were undercorrected in ⫺0.70 D compared to those from groups A (⫺0.48 ⫾ 0.60 D) and B (⫺0.44 ⫾ 0.68 D) (one-way ANOVA, P ⫽ 0.020). Moreover, significant correlations were found between refraction at 6 and 12 months after retreatment and the spherical equivalent before retreatment (r ⫽ 0.29, P ⫽ 0.035 at 6 months; r ⫽ 0.28, P ⫽ 0.037 at 12 months) in such a way that the higher the myopia before retreatment, the higher the myopia after retreatment (Fig 2). Ablation diameters at the time of retreatment were also related to 12-month postretreatment refraction, since those eyes with 5.0-mm ablation diameter were more undercorrected (⫺1.17 ⫾ 1.20 D) than those with 5.5 (⫺0.53 ⫾ 0.63 D) or 6 mm (⫺0.42 ⫾ 0.60 D) (one-way ANOVA, P ⫽ 0.021). No significant differences in regression of effect were observed between eyes operated on at 3 months after the primary LASIK and those operated on at 6 months (Mann–Whitney test, P ⫽ 0.100 regression between 3– 6 months; P ⫽ 0.776 between 6 –12 months; P ⫽ 0.133 between 3–12 months). Moreover, no differences in regression of effect were found among the myopic groups before the initial treatment (groups A, B, and C) (Kruskall–Wallis test, P ⫽ 0.182 for the regression between 3– 6 months; P ⫽ 0.851 between 6 –12 months; P ⫽ 0.512 between 3–12 months). A relationship between regression and preretreatment refraction was not present either (r ⫽ ⫺0.096, P ⫽ 0.526 for the regression between 3– 6 months; r ⫽ 0.008, P ⫽ 0.956 between 6 –12 months; r ⫽ ⫺0.205, P ⫽ 0.148 between 3–12 months).

Astigmatism Figure 2. Scattergram of attempted correction vs. refractive outcome (spherical equivalent) 12 months after retreatment.

Forty-nine percent of eyes (29 of 59) underwent a spherical laser ablation and 51% of eyes (30 of 59) an astigmatic ablation. In the spherical ablation group, mean astigmatism was 0.49 ⫾ 0.26 D

23

Ophthalmology Volume 106, Number 1, January 1999

Figure 4. Time course of mean central keratometric values (mean ⫾ standard deviation). The numbers in parentheses indicate number of eyes studied at each period.

Figure 6. Time course of corneal pachymetry (mean ⫾ standard deviation). Numbers in parentheses indicate number of eyes studied at each period.

Videokeratography (range, 0 – 0.75 D) at the time of retreatment and 0.59 ⫾ 0.36 D (range, 0 –1.00 D) at 3 months, 0.52 ⫾ 0.34 D (range, 0 –1.50 D) at 6 months, and 0.46 ⫾ 0.39 D (range, 0 –1.50 D) at 12 months after retreatment. The differences between values before and after retreatment were not significant at any follow-up (two-way ANOVA, P ⫽ 0.476) nor were the differences between values after retreatment (two-way ANOVA, P ⫽ 0.237). No eye had an astigmatism greater than 1.00 D before retreatment and only one eye (3.5%) at 12 months after retreatment. In the astigmatic ablation group, mean astigmatism was 1.25 ⫾ 0.51 D (range, 1–3.50 D) at the time of retreatment and 0.49 ⫾ 0.36 D (range, 0 –1.50 D) at 3 months, 0.54 ⫾ 0.39 D (range, 0 –1.50 D) at 6 months, and 0.52 ⫾ 0.39 D (range, 0 –1.25 D) at 12 months after retreatment. The differences between values before and after retreatment were significant for all follow-ups (two-way ANOVA, P ⬍ 0.001); however, no significant differences were found between values after retreatment (two-way ANOVA, P ⫽ 0.536). Thirty-three percent of eyes (10 of 30) had an astigmatism greater than 1.00 D before retreatment, whereas only one eye (3.7%) had an astigmatism greater than 1.00 D 12 months after retreatment.

Mean central corneal keratometric values are shown in Figure 4. These values showed marked flattening after retreatment at all follow-ups (two-way ANOVA, P ⬍ 0.001), with steepening documented (0.15 D) from the 3-month to the 6-month reading (twoway ANOVA, P ⫽ 0.040) and also (0.21 D) from the 6-month examination to the 12-month reading (P ⬍ 0.001). The corneal steepening between 3 and 12 months was 0.37 D (P ⬍ 0.001). There was a highly significant correlation between achieved correction (difference between refractions before and after retreatment) and achieved corneal flattening (difference between mean keratometric readings before and after retreatment) at 3 months (r ⫽ 0.65, P ⬍ 0.001), 6 months (r ⫽ 0.65, P ⬍ 0.001), and 12 months (r ⫽ 0.65, P ⬍ 0.001) (Fig 5). The correlation between corneal steepening and regression of refraction after retreatment was significant between 3 and 6 months (r ⫽ 0.52, P ⫽ 0.003); the greater the corneal steepening, the higher the regression of refraction. However, no correlation was found between 6 and 12 months (P ⫽ 0.138).

Pachymetry Ultrasonic pachymetric readings after retreatment were significantly lower than before retreatment (two-way ANOVA, P ⬍ 0.001) (Fig 6). Corneal thickness increased (6.2 ⫾ 8.6 ␮m) significantly between 3 and 6 months (P ⬍ 0.001), whereas no change occurred between 6 and 12 months (P ⫽ 0.237).

Intraocular Pressure Mean intraocular pressure was 12.2 ⫾ 2.1 mmHg before retreatment, and after retreatment, it was 11.7 ⫾ 1.8 mmHg at 3 months, 11.8 ⫾ 1.6 mmHg at 6 months, and 12.1 ⫾ 2.1 mmHg at 12 months. No significant change was found between values before and after retreatment (Wilcoxon test, P ⫽ 0.052).

Complications

Figure 5. Relationship between achieved correction and corneal flattening 12 months after retreatment.

24

An undercorrection greater than 1.00 D was present in 3.7% of eyes (2 of 54) at 3 months after retreatment, 14% (7 of 50) at 6 months, and 18.1% (10 of 55) at 12 months. An overcorrection greater than 1.00 D occurred in 3.7% (2 of 54), 4% (2 of 50), and

Pe´rez–Santonja et al 䡠 Retreatment after LASIK

Figure 7. Corneal flap melting on an epithelial ingrowth area 12 months after laser in situ keratomileusis retreatment.

0% (0 of 55) of eyes at 3, 6, and 12 months after retreatment, respectively. In 1.7% of eyes (1 of 59) at the time of retreatment, fine wrinkles (striations) resembling fingerprint lines were observed in the flap under broad illumination. After retreatment, 5.5% (3 of 54), 6% (3 of 50), and 5.4% (3 of 55) of eyes at 3, 6, and 12 months, respectively, had fine wrinkles. The wrinkles were confined to the flap, did not affect visual acuity, and remained unchanged during follow-up. No flap displacement was found after retreatment. Haze or scarring in the stromal bed was minimal both before and after retreatment. On a subjective scale (0 ⫽ none, 4 ⫽ severe), the mean haze rating was 0.10 ⫾ 0.2 before retreatment and 0.2 ⫾ 0.2 at 3 months, 0.1 ⫾ 0.2 at 6 months, and 0.0 ⫾ 0.0 at 12 months after reoperation. Haze was never worse than mild throughout the follow-up. Dot remnants, related to epithelial or tear film debris or to foreign particles, were minimally observed after retreatment. One month after LASIK or after LASIK enhancement, it is usual to observe an incomplete gray line at the edge of the flap due to the interaction between the epithelium and the stroma; however, in a few eyes, some epithelial ingrowth at the edge of the flap was observed. In our study, this ingrowth was recorded when it was greater than 0.5 mm in depth. Mean epithelial ingrowth at the time of retreatment was 4.2 ⫾ 16°, and after retreatment, it was 12 ⫾ 20° at 6 months and 13.5 ⫾ 24° at 12 months. There was a significant increase in epithelial ingrowth after surgery at all follow-ups (two-way ANOVA, P ⬍ 0.01). At the time of retreatment, only 8.5% of eyes (5 of 59) had epithelial ingrowth, but after retreatment, 32% of eyes (16 of 50) at 6 months and 31% (17 of 55) at 12 months had epithelial ingrowth. The most common sites for epithelial ingrowth were temporal and inferior. Epithelial ingrowth was always a peripheral disorder, was never deeper than 1.5 mm, and did not affect visual acuity or corneal astigmatism. Corneal flap melt or necrosis of the flap edge was observed in one eye (1 of 59, 1.7%) before retreatment. However, flap melting was present in 8% of eyes (4 of 50) at 6 months and 10.9% (6 of 55) at 12 months after retreatment. In these eyes, the flap melting developed on an epithelial ingrowth area, and a significant relationship was found between flap melting and epithelial ingrowth (Fisher test, P ⬍ 0.001). Flap meltings were usually very small and progressed very slowly, although one eye (1.8%) showed a more rapid progression with greater flap melt (approximately 1 mm) at 12 months (Fig 7). Because the flap melting appeared on the

peripheral flap edge, visual acuity and corneal astigmatism were not affected. Thirty-nine percent of patients (23 of 58) reported night halos before retreatment. However, after retreatment, night halos decreased significantly to 26% (14 of 54) at 3 months (McNemar test, P ⫽ 0.038), 16.3% (8 of 49) at 6 months (P ⫽ 0.003), and 22% (12 of 55) at 12 months (P ⫽ 0.038). Fifty-seven percent of patients (34 of 59) reported starbursts at night at the time of retreatment. After retreatment, starbursts were reported by 53.7% of patients (29 of 54) at 3 months, 42% (21 of 50) at 6 months, and 32.7% (18 of 55) at 12 months. There was a significant decrease in starbursts at 12 months after retreatment (McNemar test, P ⫽ 0.010). At 12 months after retreatment, those eyes with 5.5- or 6-mm ablation diameters had lower incidence of halos (8.4% and 16%, respectively) than those with 5 mm (50%) (chi-square test ⫽ 7.45, P ⫽ 0.024). No relationship was found between starbursts and ablation diameters (chi-square test ⫽ 3.03, P ⫽ 0.219 at 12 months). Decentration of ablation was calculated as the distance between the center of the laser ablation and that of the visual axis using videokeratography.11 The decentration values before and 12 months after LASIK retreatment are shown in Figure 8. There was a significant improvement in decentration after retreatment (Wilcoxon test, P ⬍ 0.01). The relationship between decentration and halos was not significant (Mann–Whitney test, P ⫽ 0.107). No patient had central islands develop before or after retreatment.

Subjective Data Sixty percent of patients said that they were satisfied with the result after the primary LASIK procedure, and 60% said that they would have LASIK again. One year after retreatment, 100% of patients were satisfied with the final result, and all patients would have surgery again.

Discussion Refractive correction of high myopia is controversial, and at this time, there is not a completely satisfactory surgical procedure for its correction.12–22 Recently, excimer laserassisted keratomileusis has generated high expectations among refractive surgeons.3– 6 Excimer lasers have been used to remove tissue, both from the cap, as in traditional keratomileusis (Buratto technique),23 and from the stromal bed, as in the in situ technique (LASIK).24

Figure 8. Percentage of eyes according to decentration before and 12 months after laser in situ keratomileusis retreatment. The numbers in parentheses indicate number of eyes.

25

Ophthalmology Volume 106, Number 1, January 1999 LASIK involves lifting a corneal flap with a microkeratome and then ablating the stromal bed with an excimer laser.1,2 Preliminary studies found that LASIK offers good results for moderate and high myopia, although clinical data are still limited.1– 6,8 Concerning complications after LASIK, limited information is also available. Undercorrections, overcorrections, flap displacement, interface debris, epithelial ingrowth, flap melting, keratitis, and irregular astigmatism are complications reported after this procedure.1– 8 Undercorrection is the most common complication, and undercorrections greater than 1.00 D have been reported in 11.3% to 59% of cases.1– 6,8 Undercorrections may be managed either by performing radial keratotomy or by lifting the flap and performing another laser ablation 3 to 6 months after surgery,9,10 although limited information is available on these techniques.10 The aim of this study was to evaluate the effectiveness, predictability, and safety of LASIK retreatment for correcting residual myopia by lifting the flap and reablating the stromal bed. Uncorrected visual acuity is the main criterion used to assess the effectiveness of a refractive procedure.25 Our study shows an improvement in UCVA after LASIK retreatment. Twelve months after retreatment, 61.8% of eyes had a UCVA of 0.5 (20/40) or better. This result is clearly better than in our own series6 of single-LASIK patients with high myopia (ⱖ ⫺8.00 D), in which 46.4% of the eyes had a visual acuity greater than or equal to 0.5 6 months after surgery. After reoperation, BCVA improved by half a line (0.054 at 12 months) over the values before retreatment. In addition, BCVA also improved after the primary LASIK procedure (0.03) in such a way that the global improvement in BCVA is nearly one line over the values before the primary LASIK procedure. These results are similar to those reported after single-LASIK for high myopia.1,5,6 Improved visual acuity is common with other refractive surgery methods,21,22 and it is the result of an enlarged retinal image.26 Concerning predictability, the mean refraction after retreatment was not as good as that obtained in eyes with only one LASIK procedure for high myopia (mean refraction 6 months after retreatment ⫺0.44 ⫾ 0.80 D, and 6 months after single-LASIK ⫹0.18 ⫾ 1.66 D6). At 6 months after retreatment, 82% of eyes were within 1.00 D of emmetropia and 81.8% at 12 months. This is a better result than in eyes that underwent only one LASIK for correcting myopia greater than ⫺8.00 D6 (in which from 46%–72% were within 1.00 D 6 months after surgery), but it is a poorer result than in eyes with only one LASIK procedure for correcting myopia lower than ⫺6.00 D8,27 (from 93%– 94.4% of eyes were within 1.00 D 6 months after surgery). Both the original refraction before the primary LASIK procedure and the preretreatment refraction served as indicators of potential successful refractive outcome after reoperation. Those eyes with an original refraction from ⫺16.00 to ⫺20.00 D were more undercorrected after retreatment than those with an original refraction from ⫺8.00 to ⫺16.00 D. In addition, the higher the myopia preretreatment, the higher the myopia after reoperation. Those eyes with 5.0-mm ablation diameter during LASIK enhancement were

26

also more undercorrected than those with wider ablations. Conversely, no significant differences were found in postretreatment refraction between eyes operated on at 3 months after the primary LASIK and those operated on at 6 months. Therefore, the LASIK algorithms used in this study should be modified, taking into account, at least, the influence of the original and the preretreatment refractions and also the ablation diameter used. In terms of stability, our study shows a regression of effect (0.21 D) between 3 and 6 months after retreatment and also between 6 and 12 months (0.17 D). For singleLASIK patients,6 a mean regression of 0.15 D was found between 3 and 6 months after surgery for preoperative myopia from ⫺8 to ⫺12 D and 0.38 D for preoperative myopia from ⫺12 to ⫺16 D, these results being similar to ours. The regression after retreatment was neither related to the myopic groups before the initial treatment nor to the preretreatment refraction. Pallikaris and Siganos1 reported that postoperative refractive astigmatism is minimally influenced by the surgical creation of a corneal flap in LASIK. In this study, no differences were found between refractive astigmatism before and after retreatment in the spherical ablation group, proving that lifting the corneal flap does not affect refractive astigmatism. In the astigmatic ablation group, refractive astigmatism decreased after retreatment, suggesting that LASIK retreatment may also be useful for correcting residual astigmatism. Some studies6,28 reported considerable early corneal flattening after LASIK followed by corneal steepening that decreases with time. This corneal steepening was related to preoperative myopia and to postoperative regression. Our keratographic results also showed a marked early flattening after retreatment followed by a significant corneal steepening of minor magnitude. The corneal steepening after LASIK retreatment (0.15 D from 3– 6 months) was lower than that reported after single-LASIK (0.23 D from 3– 6 months).6 Our keratographic study also found a strong relationship between achieved correction and achieved corneal flattening. This allows us to predict achieved correction based on the induced corneal flattening. Moreover, a significant relationship was found between regression of refractive effect and postretreatment corneal steepening between the third and sixth month. This finding suggests that regression of effect is caused by the postoperative corneal steepening, at least partially, because it was previously reported after single-LASIK.6 In our main series concerning single-LASIK,6 a significant increase in corneal thickness was observed between 1 and 3 months after surgery; thereafter, corneal thickness remained stable. In this study, a significant increase in corneal thickness was also present from 3 to 6 months with no change thereafter. These findings suggest that some changes are present in the corneal epithelium, interface, or stroma. Recently, Arbelaez and coauthors27 reported a relationship between regression after LASIK and epithelial hyperplasia. Several complications have been reported after LASIK. Undercorrection is the most common complication,1– 6,8 and in our whole group of 393 LASIK procedures, 17.5% of

Pe´rez–Santonja et al 䡠 Retreatment after LASIK eyes were undercorrected by 1.00 D or more at 3 or 6 months after surgery. After retreatment, an undercorrection greater than 1.00 D was found in 14% of eyes at 6 months and 18.1% at 12 months. Therefore, undercorrections were more common after single-LASIK than after LASIK retreatment if we compare the first 6 months. Salah and coauthors8 reported that 5.6% of eyes were overcorrected more than 2.10 D after single-LASIK, and Pe´rez–Santonja and coauthors6 found an overcorrection greater than 2.50 D in 8.6% of eyes. After LASIK retreatment, only 4% of eyes at 6 months and 0% at 12 months were overcorrected more than 1.00 D. Haze or scarring in the interface is minimal or absent both after LASIK1,5,6,8 and after LASIK retreatment. Dot remnants in the interface are rare after LASIK5,6 and also after LASIK enhancement. Dot remnants have no clinical significance and are avoided with careful cleaning during flap replacement. Fine wrinkles in the flap have been reported in 3.4% of eyes after single-LASIK,6 and they were related to flap folds during surgery.6,29 In this study, fine wrinkles were present in 5.4% of eyes at 12 months after retreatment. The wrinkles did not affect visual acuity and remained unchanged during follow-up. Epithelium in the interface is not a rare condition after LASIK (2%5– 8.8%6), although it is usually limited to the edge of the flap, does not usually progress, and does not affect visual acuity. However, epithelial ingrowth may lead to a more serious complication, flap melting or necrosis of the flap.6 Our results show that epithelial ingrowth worsens after LASIK retreatment. We believe that this increase in epithelial ingrowth is related to surgical technique. The epithelium debridement and the interface dissection with a flat spatula create an irregular epithelial border with epithelial defects, which facilitates epithelial ingrowth beneath the flap edge after surgery. Although the proper management of epithelial ingrowth is still poorly understood, we believe that some measures might decrease its incidence after LASIK enhancement, such as a more meticulous surgical technique with a linear epithelial dissection without epithelial defects, a copious irrigation of the interface for removing all implanted epithelial cells, and a strong adhesion with a minimal gap between the flap edge and the stromal bed to block epithelial ingrowth. Flap melting has been reported in 5.7% of eyes after single-LASIK.6 Flap melting usually develops over an area of epithelial ingrowth and usually shows no progression. As epithelial ingrowth, it is a peripheral disorder and does not affect visual acuity. In our study, flap melting was only observed in one eye (1.7%) before retreatment; however, it was present in 10.9% of eyes at 12 months after retreatment. These meltings were usually very small and did not progress, although one eye showed a more rapid progression. Visual acuity was not affected in any case. A relationship was also found between flap melting and epithelial ingrowth as reported for single-LASIK. Therefore, the same strategies described to decrease epithelial ingrowth would be useful to prevent flap melting. Decentration greater than 0.75 mm was reported in 6.5% of eyes after single-LASIK.6 In the current study, 12.2% of

eyes had a decentration greater than 0.75 mm at the time of retreatment. However, after retreatment, only 1.8% of eyes were decentered by more than 0.75 mm. Therefore, a remarkable improvement in decentration is observed after LASIK enhancement. Night halos and starbursts improved significantly after retreatment. At 12 months, 22% and 32.7% of patients reported night halos and starbursts, respectively, and similar results were found 6 months after single-LASIK.5,6 These night-vision symptoms improved with time, and the night halos were related to small ablation diameters. In conclusion, LASIK retreatment by lifting the flap and reablating the stromal bed is an effective and predictable procedure for correcting residual myopia after LASIK. By taking into account both the original refraction before the primary LASIK and the preretreatment refraction, we could improve the predictability of LASIK retreatment. Refraction was stable after 3 months with only minor changes. Although vision-threatening complications are rare after LASIK enhancement, epithelial ingrowth and flap melting were more common than before retreatment. On the contrary, LASIK retreatment improved decentration and night-vision problems. Although LASIK retreatment is a safe procedure, a more meticulous surgical technique with less epithelial damage would improve its safety.

References 1. Pallikaris IG, Siganos DS. Excimer laser in situ keratomileusis and photorefractive keratectomy for correction of high myopia. J Refract Corneal Surg 1994;10:498 –510. 2. Bas AM, Onnis R. Excimer laser in situ keratomileusis for myopia. J Refract Surg 1995;11(3 Suppl):S229 –S233. 3. Fiander DC, Tayfour F. Excimer laser in situ keratomileusis in 124 myopic eyes. J Refract Surg 1995;11(3 Suppl):S234 – S238. 4. Kremer FB, Dufek M. Excimer laser in situ keratomileusis. J Refract Surg 1995;11(3 Suppl):S244 –S247. 5. Gu¨ell JL, Muller A. Laser in situ keratomileusis (LASIK) for myopia from ⫺7 to ⫺18 diopters. J Refract Surg 1996;12: 222– 8. 6. Pe´rez–Santonja JJ, Bellot J, Claramonte P, et al. Laser in situ keratomileusis to correct high myopia. J Cataract Refract Surg 1997;23:372– 85. 7. Pe´rez–Santonja JJ, Sakla HF, Abad JL, et al. Nocardial keratitis after laser in situ keratomileusis. J Refract Surg 1997;13: 314 –7. 8. Salah T, Waring GO 3rd, el Maghraby A, et al. Excimer laser in situ keratomileusis under a corneal flap for myopia of 2 to 20 diopters. Am J Ophthalmol 1996;121:143–55. 9. Hollis S, Rozakis GW. Complications, special cases and management. In: Rozakis GW, ed. Refractive Lamellar Keratoplasty. Thorofare, NJ: SLACK Inc., 1994;111–22. 10. Machat JJ, ed. Excimer laser refractive surgery: practice and principles. Thorofare, NJ: SLACK, 1996;403–13. 11. Cavanaugh TB, Durrie DS, Riedel SM, et al. Topographical analysis of the centration of excimer laser photorefractive keratectomy. J Cataract Refract Surg 1993;19(Suppl):136 – 43. 12. Bas AM, Nano HD Jr. In situ myopic keratomileusis, results in 30 eyes at 15 months. Refract Corneal Surg 1991;7:223–31.

27

Ophthalmology Volume 106, Number 1, January 1999 13. Kim WJ, Lee JH. Long-term results of myopic epikeratoplasty. J Cataract Refract Surg 1993;19:352–5. 14. Goosey JD, Prager TC, Goosey CB, Martin DI. One year follow-up of epikeratoplasty for myopia. J Cataract Refract Surg 1990;16:21–30. 15. Werblin TP, Patel AS, Barraquer JI. Initial human experience with Permalens myopic hydrogel intracorneal lens implants. Refract Corneal Surg 1992;8:23– 6. 16. Barraquer C, Cavelier C, Mejı´a LF. Incidence of retinal detachment following clear-lens extraction in myopic patients. Retrospective analysis. Arch Ophthalmol 1994;112: 336 –9. 17. Colin J, Robinet A. Clear lensectomy and implantation of low-power posterior chamber intraocular lens for the correction of high myopia. Ophthalmology 1994;101:107–12. 18. Maxwell WA. Myopic keratomileusis: initial results and myopic keratomileusis combined with other procedures. J Cataract Refract Surg 1987;13:518 –24. 19. Keratophakia and keratomileusis: safety and effectiveness. American Academy of Ophthalmology. Ophthalmology 1992; 99:1332– 41. 20. Heitzmann J, Binder PS, Kassar BS, Nordan LT. The correction of high myopia using the excimer laser. Arch Ophthalmol 1993;111:1627–34.

21. Fechner PU, Strobel J, Wicchmann W. Correction of myopia by implantation of a concave Worst-iris claw lens into phakic eyes. Refract Corneal Surg 1991;7:286 –98. 22. Pe´rez–Santonja JJ, Bueno JL, Zato MA. Surgical correction of high myopia in phakic eyes with Worst–Fechner myopia intraocular lenses. J Refract Surg 1997;13:268 – 81. 23. Buratto L, Ferrari M, Rama P. Excimer laser intrastromal keratomileusis. Am J Ophthalmol 1992;113:291–5. 24. Pallikaris IG, Papatzanaki ME, Stathi EZ, et al. Laser in situ keratomileusis. Lasers Surg Med 1990;10:463– 8. 25. Lynn MJ, Waring GO III, Carter JT. Combining refractive error and uncorrected visual acuity to assess the effectiveness of refractive corneal surgery. Refract Corneal Surg 1990;6:103–9. 26. Applegate RA, Howland HC. Magnification and visual acuity in refractive surgery. Arch Ophthalmol 1993;111:1335– 42. 27. Arbelaez MC, Pe´rez-Santonja JJ, Ismail MM, et al. Automated lamellar keratoplasty (ALK) and laser in situ keratomileusis (LASIK). In: Serdarevic ON, ed. Refractive Surgery: Current Techniques and Management. New York: Igaku-Shoin, 1997; chap 9. 28. Pallikaris IG, Siganos DS. Laser in situ keratomileusis to treat myopia: early experience. J Cataract Refract Surg 1997;23:39–49. 29. Gimbel HV, Basti S, Kaye GB, Ferensowicz M. Experience during the learning curve of laser in situ keratomileusis. J Cataract Refract Surg 1996;22:542–50.

Discussion by Mark E. Whitten, MD The excimer laser has been found to be a safe and effective tool for the correction of myopia. Haze and/or scarring of the cornea occurred as correction of high myopia was attempted with surface ablation. Laser in situ keratomileusis (LASIK) has been offered as an alternative to surface photorefractive keratectomy (PRK) for the correction of high myopia.1,2 The current study examines the effectiveness, predictability, and complications of retreatment of LASIK patients. It is interesting to note in the study that only 15% of 393 eyes with high myopia greater than ⫺8.00 diopters (D) required retreatment. This low retreatment rate is particularly encouraging in this highly myopic group. The retreatment was shown to be effective and predictable with significant improvement in uncorrected visual acuity and best-corrected visual acuity. They found that the higher the preretreatment myopia, the greater the myopia after retreatment. A correlation is established between regression and increased corneal steepening and thickening. They also found retreating astigmatism was very effective while there was no effect on astigmatism if the flap was lifted to only retreat residual myopia. Reviewing their complications, it is gratifying to note that lasting overcorrection did not occur. Haze or scarring was not an issue. Visually significant folds in the flap were not present after lifting. Night halos and starbursts actually improved with retreatment and enlargement of optical zones. The incidence of epithelial ingrowth (31%) in the retreatment group is disturbing. The rate of flap melting was 10.9%. The authors note that visual acuity was not affected by these occurrences. They believe this increase in epithelial ingrowth

From the Department of Keratorefractive Service, Washington National Eye Center, Washington Hospital Center, 8218 Wisconsin Avenue, Suite #104, Bethesda, MD 20814.

28

over that found in primary LASIK was related to their surgical technique. A procedure should be performed only if the benefits outweigh the risks. The authors have clearly shown that the benefits of retreatment are significant with greatly improved visual acuity and lessening of irritating night halos and starbursts. A flap melting rate of 10.9% with an associated epithelial ingrowth rate of 31% is concerning. Epithelial ingrowth should only be treated when vision is threatened or when flap melting occurs. Vision was not threatened in these cases. Flap melting does represent a real risk to the patient for vision loss. Patients with epithelial ingrowth need to be watched carefully to prove that flap melting will not occur. If the authors find that this rate of ingrowth cannot be modified by a change in technique, they may wish to substitute recutting the flaps 4 months after the primary LASIK so that the incidence of epithelial ingrowth then would approach that of the primary procedure. This excellent retrospective study allows us to hope that a prospective study will show improved predictability based on preretreatment myopia, pachymetry, and topography. Videotape recording of retreatment cases may help establish better techniques to avoid epithelial ingrowth. As the development of new procedures moves from the traditional academic setting to private practitioners, it becomes incumbent on us all to study our patients to improve our outcomes. With thousands of patients undergoing PRK, LASIK, and retreatment, it would be disappointing to lose the opportunity to analyze these patients in a meaningful way. References 1. Machat JJ. Excimer laser refractive surgery. Practice and Principles. Thorofare: Slack Inc, 1996. 2. Kremer FB, Dufek M. Excimer laser in situ keratomileusis. J Refract Surg 1995;11:S244 –S247.