Corneal Topography of Excimer Laser Photorefractive Keratectomy Using a 6-mm Beam Diameter

Corneal Topography of Excimer Laser Photorefractive Keratectomy Using a 6--mm Beam Diameter Peter S. Hersh, MD,l Shetal1. Shah, AB,2 Summit PRK Topography Study Group* Objective: The purpose of the study is to define qualitative pattems of comeal topography after excimer laser photo refractive keratectomy (PRK) using a 6-mm beam diameter, investigate changes in pattems over time, and identify associations of topography pattems with clinical outcomes. Design: Multicenter, prospective cohort study. Participants: Ninety-eight eyes of 90 patients with myopia who had undergone PRK using the Summit Technology, Inc., excimer laser with a 6-mm beam diameter. Intervention: Computer-assisted videokeratography data were analyzed for eyes having undergone PRK. Topography pattems at 3, 6, and 12 months after surgery were classified and associations with clinical outcomes assessed. Main Outcomes Measured: Topography pattems after PRK were determined at 3, 6, and 12 months after surgery. Associations with preoperative characteristics of age and attempted correction, and postoperative outcomes of uncorrected and spectacle-corrected visual acuity, predictability, astigmatism, comeal haze, glare, halo, and patient satisfaction were analyzed. Results: At 1 year, 21.4% of comeas showed a homogeneous topography, 27.6% showed a toric-with-axis configuration, 10.2% showed a toric-against-axis configuration, 7.1 % showed an irregularly irregular topography, 24.5% showed a keyhole/semicircular pattem, and 9.2% showed focal topographic variants. From 3 to 6 months, 40.1 % of maps changed; from 6 to 12 months, 53.1 % of maps changed, generally to optically smoother, regular pattems. Older age and higher attempted correction were associated with the development of more irregular pattems. The irregular groups showed worse predictability than did the regular groups and a tendency for slight overcorrection. The average reported glare/halo of 1.33 (scale = 0 to 5) in this study was less than in a previous study of the 4.5- to 5-mm treatment zone. However, of six patients expressing dissatisfaction with the results of surgery, three ranked their glare or halo at the maximum level. Conclusions: Topography pattems using a 6-mm beam diameter are identifiable, improve with time, and may affect clinical outcomes after photorefractive keratectomy (PRK). The keyhole/semicircular pattem is more prevalent with a 6 mm treatment zone than with smaller treatment zones. Although optical side effects of glare and halo appear to be reduced with the 6-mm treatment, a small number of patients still report substantial glare or halo after the procedure. Ophthalmology 1997;104:1333-1342 Originally received: July 23, 1996. Revision accepted: January 27, 1997. 1 Department of Ophthalmology, UMDNJ-New Jersey Medical School, Newark, New Jersey and the Cornea and Laser Institute, Hackensack University Medical Center, Teaneck, New Jersey. Department of Ecology and Evolutionary Biology, Princeton University, Princeton, New Jersey.

2

* Members of the Summit PRK Topography Study Group are listed in the Appendix at the end of this article. Presented in part at the Annual Meeting of the American Academy of Ophthalmology, Chicago, November 1996.

Recently, the U.S. Food and Drug Administration approved the use of the excimer laser for the correction of mild-to-moderate myopia using a 6-mm beam diameter. Supported in part by an unrestricted grant to the Department of Ophthalmology from Research to Prevent Blindness, Inc, New York; and Summit Technology, Inc, Waltham, Massachusetts. Dr. Hersh is a consultant for Summit Technology, Inc. Reprint requests to Peter S. Hersh, MD, Department of Ophthalmology, UMDNJ-New Jersey Medical School, 90 Bergen Street, 6th Floor, Newark, NJ 07103.

1333

Ophthalmology

Volume 104, Number 8, August 1997

Because this larger beam diameter may be associated with a postoperative corneal topography different from that which results from the smaller treatment zone size, it is important to assess the characteristics of corneal topography after 6-mm PRK. In this article, we analyze the corneal topography of the 6-mm treatment zone using the Summit Technology excimer laser (Waltham, MA) and assess clinical associations.

Patients and Methods Study Design and Photorefractive Keratectomy Procedure As part of ongoing multicenter clinical studies of the Summit Technology, Inc, excimer laser (Waltham, MA), the corneal topography of 98 eyes of 90 patients who underwent photorefractive keratectomy with a 6-mm beam diameter was investigated. Attempted corrections ranged from 1.50 to 6.00 diopters (D), and all patients in the study had less than or equal to 1.00 D of regular astigmatism. Procedures were performed at one of five separate centers. All study centers conformed to standardized patient entry criteria under an Investigational Device Exemption granted to Summit Technology, Inc. Preoperative and follow-up visits included a detailed ophthalmologic examination, visual acuity under controlled lighting conditions using an Early Treatment Diabetic Retinopathy Study chart, and computer-assisted videokeratography (EyeSys Laboratories, Houston, TX). Corneal haze was determined during slit-lamp examination and reported on a scale of 0 to 4. Photorefractive keratectomy was performed with the Summit Technology OmniMed excimer laser system using a beam diameter of 6 mm. In all cases, the procedure was centered over the entrance pupil with the patient coaxially fixating, as suggested by Uozato and Guyton. I - 3 To ensure appropriate laser energy and beam homogeneity, ablation and beam profile characteristics were tested at the beginning of each treatment day. This and the actual PRK procedure are detailed elsewhere. 4 After surgery, a combination antibiotic-steroid ointment and a patch were applied. Per the study protocol, the ointment was continued five times daily until the cornea had re-epithelialized. Fluorometholone 0.1 % drops then were administered four times daily for 1 month, three times daily for 1 month, two times daily for 2 weeks, once daily for 1 week, and then every other day for 1 week before discontinuation. Subsequent corticosteroid use was at the individual surgeon's discretion. Analysis of Corneal Topography Data Acquisition. Videokeratography was obtained at each study site by a trained technician with the EyeSys Corneal Analysis System (EyeSys Laboratories, Houston, TX). Subsequent data analysis was performed by two investigators (PSH, SIS). In this report, topography data

1334

taken before surgery, and 3 months, 6 months, and 1 year after surgery were analyzed. Characterization and Clinical Effects of Treatment Zone Topography. The qualitative changes in corneal topography produced by the PRK procedure were determined by review of the differential topography map, derived from subtraction of the preoperative from postoperative power at corresponding points on the topography system's axial power maps. The differential map was used rather than the postoperative power map, because it best reflects the changes in corneal contour, which can be ascribed to the laser treatment itself and subsequent wound healing. Color bins of 0.5 D were used for all analyses. After a preliminary review did not show newly definable patterns with the 6-mm treatment zone, each color map was allocated to one of seven defined treatment zone topography categories. These patterns are illustrated in previous work4 : 1. Homogeneous, which shows a uniform and symmetric flattening. 2. Toric-with-axis, which shows a meridionally symmetric treatment zone with a greater induced flattening in the steep preoperative axis. 3. Toric-against-axis, which shows a meridionally symmetric treatment zone with a greater induced flattening in the flat preoperative axis. 4. Irregularly irregular, which shows generalized irregularities over the treatment zone, defined as more than one area measuring greater than 0.5 mm and greater than 0.5 D in power from other areas at the same radius from the optical zone center, or 1 area measuring greater than 1.0 mm in size and 1.0 D in power not conforming to the specific criteria of any other patterns described. 5. Keyhole/semicircular, which shows topographic regions, quantitatively measuring greater than 1.0 mm in size and 1.0 D of relatively less flattening, extending in from the periphery of the ablation zone (keyhole), or a general foreshortening of the ablation zone effect in 1 meridian (semicircular). 6. Central island, which shows a central area of relatively less flattening measuring greater than 1.0 mm in size and greater than 1.0 D in power and not extending to the periphery. 7. Focal topographic variant, which shows a generally homogeneous pattern with irregularities measuring less than 1.0 mm in size or less than 1.0 D in power. All maps were graded by two masked observers. Any differences were resolved subsequently by joint review. In addition, all qualitative characteristics of the color maps were verified by review of the actual numeric data used to derive the topography maps. Only one map was available for each subject. Topography maps were available for 98 patients at all 3 timepoints. Descriptive statistics were generated for the seven topography groups. The topography patterns were tested for relations with preoperative characteristics of patient age and attempted correction and clinical outcomes of (1) uncorrected visual acuity, (2) spectacle-cor-

Hersh et al . Corneal Topography of Excimer Laser PRK rected visual acuity, (3) predictability (achieved-attempted correction), (4) change in refractive astigmatism, (5) glare and halo, (6) subjective patient satisfaction, and (7) corneal haze. Mean visual acuities were derived by the method of Holladay and Prager5 using log of the minimum angle of resolution (LogMAR) acuities. On self-administered subjective patient questionnaire forms , patients ranked both glare and halo from 0 to 5 where 5 represented the maximum symptom. Patient satisfaction similarly was determined by individual patient subjective grading of 0 to 5. Corneal haze was ranked on slit-lamp biomicroscopy on a scale of 0 to 4. For greater power on statistical analysis, the topography subgroups also were collapsed into two broader categories-the homogeneous, toric-with-axis, and toricagainst-axis were combined into one broader regular group and the irregularly irregular, keyhole/semicircular, central island, and focal topographic variants were combined into one broader irregular group-and similarly analyzed. In previous work, we have validated the selection of these patterns as regular or irregular using a quantitative descriptor of corneal topography.6 Statistical Analysis. The LogMAR values for uncorrected and best-corrected visual acuity and predictability measures were approximately normally distributed. Therefore, differences in these outcomes among topography groups were compared by analysis of variance (ANOVA) (for the multiple group comparison) and t tests (for the comparison of pooled regular and irregular groups). The changes in astigmatism were calculated by taking the nominal difference between the preoperative and postoperative astigmatism for each eye. Analysis of variance then was used to identify whether there were any differences in astigmatism change among the different topography groups. Chi-square tests were used to analyze the outcomes of glare, halo, satisfaction, and corneal haze. As listed in the tables, the number of patients included in each analysis excludes those persons for whom the specific data were unavailable.

Results Characterization of Treatment Zone Topography and Changes Over Time Table 1 and Figure 1 show the topography classification at 3, 6, and 12 months after PRK. At 3 and 6 months, the keyhole/semicircular group comprised the greatest number of patients (36.7% and 32.7%, respectively). At 1 year, the toric-with-axis group comprised the highest percentage (27.6%). Table 2 lists the change in topography patterns over time. Comparing the 3- and 6-month examination results, 40 (40.1 %) of98 eyes had a change in pattern. This period saw the percentage of eyes in the pooled regular group increase from 44.9% to 46.9%, whereas the percentage in the pooled irregular group decreased correspondingly fro m 55.1% to 53.1%. To further analyze these changes, patterns were grouped according to their presumed clinical desirability:

homogeneous and toric-with-axis greater than toricagainst-axis and focal topographic variant greater than keyhole/semicircular and irregularly irregular. Using this schema, of those 40 eyes with a change in topography pattern from month 3 to 6, 18 (45 %) changed to a clinically better pattern, 9 (22.5%) remained in an equivalent category, and 13 (32.5 %) worsened. Comparing the 6-month and I-year examination results, 52 (53.1 %) of 98 eyes had a change in topography pattern. This period saw the percentage of eyes in the pooled regular group increase from 46.9% to 59.2%, whereas the percentage in the pooled irregular group decreased correspondingly from 53 .1 %to 40.8%. In particular, the keyhole/semicircular group decreased from 32.7% to 24.5%, whereas the percentage of eyes in the homogeneous and toric-with-axis groups increased. For those 52 eyes with a change in topography pattern from month 6 to 12,25 (48%) changed to a clinically better pattern, 14 (27%) remained in an equivalent category, and 13 (25%) worsened. Preoperative Variables Age. At 3 months, the average age in the combined regular group was 35.4 (standard deviation [SD] = 7.4) compared with 36.1 (SD = 7.8) in the combined irregular group. At 6 months, the average age in the regular group was 34.9 (SD = 6.9) compared with 36.5 (SD = 8.2) in the irregular group. At 1 year, the average age in the regular group was 35.9 (SD = 7.6) compared with 35.6 (SD = 7.7) in the irregular group. There was no difference in mean age between the pooled regular and irregular groups at each timepoint using t tests. When results were stratified by decade of life, however, there was a statistically significant difference at 6 months with a tendency for the irregular group to be somewhat older (P = 0.016, chi-square test with 2 degrees of freedom). Attempted Correction. At all three timepoints, the average attempted correction was found to be greater in the combined irregular group than in the combined regular group (4.35 vs. 3.48 D at 3months, 4.13 vs. 3.75 D at 6 months, and 4.04 vs. 3.89 D at 1year). This was statistically significant at 3 months (P = 0.001 , t test). This finding was corroborated when eyes were stratified to 1.0D increments of attempted correction (P = 0.017, chisquare test with 5 degrees of freedom). Clinical Effects of Treatment Zone Topography Uncorrected Visual Acuity. Table 3 lists the mean uncorrected visual acuity for each topography pattern at the I-year follow-up visit. No difference was found among individual patterns (P = 0.38, ANOVA) or when comparing the two-category collapsed groups. Overall, only two eyes had an uncorrected visual acuity less than 20/40. Both had a visual acuity of 20/80 and were in the pooled irregular group. Spectacle-corrected Visual Acuity. Table 4 lists the mean best-corrected vi sual acuity for each topography pattern at the I-year follow-up visit. No difference was found among individual patterns (P = 0.40, ANOV A) or

1335

Ophthalmology

Volume 104, Number 8, August 1997

Table l. Qualitative Topography Classification (N = 98) 3 mos

1 yr

6 mos

Topography Classification

N

%

N

%

N

%

Homogeneous T oric-with-axis Toric-against-axis Irregularly irregular Keyhole/semicircular Focal topographic variant Central island Pooled regular* Pooled irregulart

11 19 14 6 36 12 0 44 54

11.2 19A 14.3 6.1 36.7 12.2 0 44.9 55.1

17 20 9 7 32

17.3 20A 9.2 7.1 32.7 13.2 0 46.9 53.1

21

21.4 27.6 10.2 7.1 24.5 9.2 0 59.2 40.8

* Pooled

13

0 46 52

27

10 7 24 9 0 58 40

homogeneous, toric-with-axis, toric-against-axis.

t Pooled irregular, keyhole/semicircular, focal topographic variant, central island.

when comparing the two-category collapsed groups. All except one eye in the study cohort had a best-spectaclecorrected visual acuity of 20120 or better. Predictability (Achieved-Attempted Refractive Correction). Table 5 lists the mean achieved-attempted manifest refractive correction at the I-year follow-up visit for each topography pattern. For all patients, disregarding topography category, the mean achieved-attempted refractive correction was +0.11 D spherical equivalent overcorrection. The homogeneous group was the most tightly grouped (SD = 0.07), whereas the irregularly irregular group showed the greatest scatter (SD = 1.09). There was no statistically significant difference in mean predictability when comparing the six categories (P = 0.28, ANOVA). Comparing the pooled groups, however, the irregular group showed a mean difference between achieved and attempted correction that was significantly different from zero (P = 0.015, t test); that is, these eyes were modestly overcorrected.

Change in Astigmatism. Table 6 lists the average change in manifest refractive cylindrical correction at 1 year after PRK stratified to topography pattern. No difference was found among topography groups or between the two pooled groups (P = 0.53, ANOVA). The irregularly irregular group showed the greatest mean change in refractive astigmatism, however. Glare and Halo. For all patients, disregarding topography category, the mean value for glare was 1.00 and for halo symptoms was 1.21 (range, 0-5). The mean glare/ha1o index (maximum grade of either glare or halo for an individual eye) was 1.33. Tables 7 and 8 list the mean values for glare and halo, respectively. There was no statistically significant correlation of topography category with either glare or halo alone or with the combined glare/halo index (P = 0.044, chisquare test with 20 degrees of freedom). Six patients at 1 year ranked glare at 4 or 5. Of these ranks, two were keyhole/semicircular, two were toricwith-axis, and one each was homogeneous and focal topo-

~r------------------------------------------------------------.

Figure 1. T opography patterns at 3, 6, and 12 months after photorefractive keratectomy in 98 eyes. Homog = homogeneous, TWA = toric-with-axis, T AA = toric-against-axis, irreg = irregularly irregular, KH/SC = keyhole/semicircular, FTV = focal topographic variam.

25

15

10

TWA

TAA

lrreg

Topography Pattern

1336

KHlSC

FTV

Hersh et al . Corneal Topography of Excimer Laser PRK Table 2. Change in Topography over Time (N = 98) 6 mos

Homogeneous

Toric-WA

Toric-AA

FTV

KH/SC

lrreg

3 mos Homogeneous Toric-WA Toric-AA ITV KH/SC Irreg

5 6 1 2 3 0

1 10 3 2 3 1

2 1 6 0 0 0

1 2 2 5 3 0

2 0 2 1 27 0

0 0 0 2 0 5

8 2 2 3 5 1

4 16 2 2 3 0

1 1 2 3 2 1

3 0 2 0 4 0

1 0 1 5 16 1

0 1 0 0 2 4

12 mos

6 mos Homogeneous Toric-WA Toric-AA FTV KH/SC Irreg

T oric-W A = toric pattern, greater flattening in steep axis; T oric-AA = toric pattern, greater flattening in flat axis; KH/SC = keyhole/semicircular; ITV = focal topographic variant; Irreg = irregularly irregular.

graphic variant. Eight patients at 1 year ranked halo at 4 or 5. Of these ranks, three were keyhole/semicircular, two were homogeneous, and one each was toric-with-axis, toric-against-axis, and focal topographic variant, respectively. Subjective Patient Satisfaction. For all patients, disregarding topography category, the mean value for patient satisfaction was 4.26 (range, 0-5). No statistically significant difference was found among the different topography patterns (P = 0.18, chi-square test with 10 degrees of freedom). At 1 year, six patients had a low satisfaction ranking of 0 or 1. Of these, three were homogeneous and one each was keyhole/semicircular, toric-with-axis, and focal topographic variant, respectively. Three of these six patients ranked their glare or halo at the maximum of 5. Corneal Haze. Table 9 lists the degree of corneal haze

Current photorefractive keratectomy treatments use a beam diameter of 6 mm. A number of clinical and theoretical studies have suggested better results in general with the 6-mm treatment zone compared to the smaller 4.5 mm or 5 mm zones. 3,7,8 Change in the size of the beam diameter of an excimer laser may be expected to influence

Table 3. Uncorrected Visual Acuity at the I-year Follow-up

Table 4. Spectacle-corrected Visual Acuity at the I-year Follow-up

stratified to topography pattern. There was no statistically significant difference among the groups (P = 0.55, chisquare test with 10 degrees of freedom). Only three patients in this study cohort showed a haze grade greater than 1. Two of these were in the keyhole/semicircular group and one was in the irregularly irregular group.

Discussion

Topography Classification

N

Mean

Standard Deviation (in Snellen lines)

Topography Classification

N

Mean

Standard Deviation (in Snellen lines)

All patients Homogeneous T oric-with-axis T oric-against-axis Irregularly irregular Keyhole/semicircular Focal topographic variant Pooled regular* Pooled irregulart

92 20 25 9 7 22 9 54 38

20/17 20/16 20/20 20/16 20/21 20/17 20/14 20/18 20/17

1.9 1.6 1.9 1.6 2.8 2.4 1.3 1.8 2.3

All patients Homogeneous T oric-with-axis Toric-against-axis Irregularly irregular Keyhole/semicircular Focal topographic variant Pooled regular* Pooled irregulart

92 20 25 9 8 22 9 55 38

20/13 20/13 20/15 20/13 20/15 20/13 20/13 20/13 20/13

1.1 0.9 1.1 1.1 1.3 1.1 0.9 1.1 1.1

* Pooled homogeneous,

toric-with axis, toric-against-axis.

t Pooled irregular, keyhole/semicircular, focal topographic variant, central island.

* Pooled homogeneous,

toric-with-axis, toric-against-axis.

t Pooled irregular, keyhole/semicircular, focal topographic variant, central

island.

1337

Ophthalmology

Volume 104, Number 8, August 1997

Table 5. Predictability (Achieved - Attempted Correction) at the I-year Follow-up Topography Classification

N

Mean

Standard Deviation

Homogeneous T oric-with-axis T oric-against-axis Focal topographic variant Keyhole/semicircular Irregularly irregular Pooled regular* Pooled irregulart

20 26 9 8 23 7 55 38

0.01 -0.05 -0.27 0.45 0.31 0.12 -0.03 0.31

0.Q7 0.85 0.89 0.45 0.67 1.09 0.80 0.72

* Pooled homogeneous,

the postoperative corneal topography in two ways. First, the macroscopic corneal optical architecture will vary with size of the treatment zone. 9•10 For instance, overall corneal asphericity will be affected l l ,12 and edge effects resulting from the junction of the ablation zone with the normal cornea should be diminished with a larger diameter treatment. 6,13,14 Second, the characteristics of the actual treatment zone topography specifically may differ with beam diameter. In past work, we have analyzed a number of characteristics of corneal topography after PRK with a beam diameter of 4.5 or 5 mm. 1,4,6,11 Because the information available from corneal topography analysis may help to explain clinical outcomes and complications and be used, it is hoped, to refine the procedure to produce a corneal optical surface that optimizes vision and minimizes optical aberrations, in this study we have investigated corneal topography after PRK using a 6-mm treatment zone. As in all studies of photorefractive keratectomy and corneal topography, it should be stressed that this investigation involved the use of one manufacturer's device.

Table 6. Change in Refractive Astigmatism at the I-year Follow-up Topography Classification Homogeneous T oric-with-axis T oric-against-axis Focal topographic variant Keyhole/semicircular Irregularly irregular Pooled regular* Pooled irregulart

N

Mean Change Astigmatism

Standard Deviation

20 26 9 8 23 7 55 38

0.02 0.03 -0.14 0.09 0.03 0.27 0.00 0.09

0.41 0.48 0.34 0.45 0.34 0.34 0.43 0.37

toric-with-axis, toric-against-axis.

t Pooled irregular, keyhole/semicircular, focal topographic variant, central

1338

Topography Classification

N

Mean

Standard Deviation

Al patients Homogeneous T oric-with-axis T oric-against-axis Irregularly irregular Keyhole/semicircular Focal topographic variant

91 20 25 9 7 22 8

1.00 0.74 0.96 1.11 1.28 1.13 1.00

1.39 1.28 1.33 1.16 1.38 1.68 1.41

0 to 5.

toric-with axis, toric-against-axis.

island.

island.

7. Glare* at the I-year Follow-up

* Subjective ordinal scale from

t Pooled irregular, keyhole/semicircular, focal topographic variant, central

* Pooled homogeneous,

Table

The findings presented, therefore, may be specific to this particular laser and not necessarily applicable to PRK in general. Also, patients in this study had 1.0 D or less of astigmatism; thus, results can not be generalized to those with more than 1.0 D of astigmatism. 1O As discussed in detail previously,4 moreover, studies such as this one need to be interpreted with a knowledge of possible errors and idiosyncrasies with current corneal videokeratography units, including the one used in this investigation. 15-20 When comparing studies of topography after PRK, the reader also should pay close attention to the specific definition of topography patterns used, because the literature, to date, variously may ascribe different definitions to a single topography pattern. Qualitative Characterization of 6,mm Topography Lin et al21 classified topography patterns using the VISX Twentyrrwenty excimer laser and a 6-mm beam diameter (VISX, Inc, Sunnyvale, CA) as central uniform (corresponding to this study's homogeneous pattern), keyhole, semicircular, and central bump (i.e., central island). At 1 month, they found an incidence of these patterns of 45%, 33%, 12%, and 10%, respectively, results comparable with 3-month findings in our study of 11.2% homogeneous, 33.7% in one of two toric patterns, 36.7% in the combined keyhole/semicircular group, and 6.1 % in the irregularly irregular group. Levin and co-workers, using a VISX Twentyrrwenty laser (VIS X , Inc) and a 6-mm treatment zone, found that

Table 8. Halo* at I-year Follow-up Topography Classification

N

Mean

Standard Deviation

All patients Homogeneous Toric-with-axis T oric-against-axis Irregularly irregular Keyhole/semicircular Focal topographic variant

91 20 25 9 7 22 8

1.21 1.15 1.24 1.22 1.57 1.08 1.40

1.45 1.66 1.23 1.56 1.27 1.65 1.41

* Subjective ordinal scale from 0 to 5.

Hersh et al . Corneal Topography of Excimer Laser PRK 67% of eyes treated had a central island pattern 3 months following the procedure. 22 These investigators defined a central island as any part of the treatment zone surrounded by areas of lesser curvature on more than 50% of its boundary, a definition more encompassing than that used in our analysis. The presence of central islands had no impact on clinical outcomes in their study, although others have ascribed patient symptoms of diplopia and glare/ halo to central islands?3 Although there were no frank central islands found as specifically defined in our investigation, it is possible in some cases that the keyhole/semicircular pattern represents a forme fruste of the central island pattern and that the two patterns have a similar etiology as discussed below. Conversely, the keyhole/semicircular pattern may represent an earlier central island that has smoothed over time with epithelial and stromal wound remodeling. The earliest topography maps analyzed in this study were taken 3 months after surgery. In a study currently in progress, we have, in fact, seen frank central islands in the earlier postoperative period that subsequently evolved to the keyhole/semicircular pattern (Hersh, 1996, unpublished data). Patterns, indeed, were found to change over time in 40.1 % of eyes from the 3- to 6-month examination and in 53.1% of eyes from the 6-month to I-year visit. With corneal epithelial and stromal healing and remodeling, changes generally were toward a smoother topography pattern. In particular, a decreased number of keyhole/ semicircular patterns was seen, whereas the homogeneous and toric patterns increased in number. This general tendency toward improved topography regularity with time, particularly with an increase in the homogeneous and toric-with-axis patterns, was similar to that found in our previous 4.5- to 5-mm study.4

Table 9. Corneal Haze* at I-year Follow-up Corneal Haze Grade Topography Classification Homogeneous T oric-with-axis T oric-against-axis Focal topographic variant Keyhole/semicircular Irregularly irregular

0

1

[N (%»)

[N (%»)

14 16 5 4 14 3

6 10 4 4 7 3

(70) (62) (56) (50) (61) (43)

(30) (38) (44) (50) (31) (43)

2-3

[N (%»)

o (0) o (0) o (0) o (0)

2 (9) 1 (14)

* Ordinal scale from 0 to 4.

Figure 2 contrasts results from this study using the 6-mrn beam to those of our previous study using the 4.5- or 5-mrn treatment zone. At 1 year, the keyhole/semicircular group comprised 24.5% of eyes with a 6-mrn treatment zone compared with 2.8% treated with the smaller diameter beam. The toric groups comprised 37.8% of eyes in this study compared with 20.5% in the study of the smaller treatment zone, whereas the percentage of eyes in the homogeneous group was less with the larger treatment zone. Potential Causes for Differences in Topography Characteristics with Different Beam Diameters Topography patterns with different beam sizes, theoretically, may differ from a number of causes as listed below. Cross-sectional differences and possible inhomogeneities in the radiant energy density profile of the laser beam between beams of different diameters. Areas of the beam with greater energy fluence will ablate relatively more

Figure 2. Topography patterns at 1 year comparing the 6-mm treatment zone (98 eyes) with the 4.5- to 5-mm treatment zone of a previous study (181 eyes) (A). Homog = homogeneous, TWA = toric-with-axis, T AA = toric-against-axis, irreg = irregularly irregular, KH/SC = keyhole/semicircular, FTV = focal topographic variant.

Homog

TWA

TAA

'"09

KHlSC

FTV

Topography Pattern

1339

Ophthalmology

Volume 104, Number 8, August 1997

Figure 3. One possible cause of topography patterns. Top, schematic depicting shifts in corneal hydration (arrows) during ablation possibly leading to, bottom left, a central island pattern or, bottom right, the keyhole/semicircular pattern. On the schematic topography maps, blue areas are flattest, orange is relatively steeper, and red is steepest (untreated area of cornea in the schematic).

Figure 4. Central island 1 week after surgery secondary to epithelial irregularity. Note (left) the irregular epithelial surface (arrow) reconstructed as (right) a central island pattern (arrow) on the color map.

tissue than those areas of less fluence. A more gaussian energy profile of the smaller diameter beam, in contrast to a "flatter" profile of a 6-mm beam, could mitigate against central island formation, because relatively more tissue would be removed at the center of the ablation zone. Similarly, the keyhole/semicircular pattern theoretically could arise from a beam that is assymetric in one meridian, causing a decrease in the corneal radius of curvature in the meridian of greater tissue ablation. Differential masking of the incoming beam by the effluent plume. 24.25 The shape and nature of the plume vortex with each pulse could differ between smaller and larger beam diameters. Moreover, the greater tissue volume removed and longer procedure times with larger diameter ablations possibly could exacerbate such problems. Differences in the laser-tissue interaction. Acoustic shockwave phenomena may differ depending on beam diameter, potentially resulting in changes in stromal hydration characteristics during the procedure. 26 Such variations in stromal hydration may lead to differential ablation rates across the treatment zone. In particular, the acoustic shockwave could drive water centrally, leading to a diminished effective stromal ablation rate in areas of the treatment zone that have become relatively overhydrated. 27 Such variations in ablation rate thus could produce a keyhole/semicircular or central island pattern (Fig 3). The shockwave and hydration phenomena may depend on three features associated with the beam diameter: the diameter itself (larger diameters causing a greater shockwave), the increased length of time of procedures using a larger beam diameter, and the energy profile of the laser. The increased incidence of the keyhole/semicircular pattern in this study compared with that of the previous study of the 5-mm beam thus may be a result of these phenomena being of greater importance with the wider beam. In addition, this hypothesis could explain the association that was found in this study between the irregular topography patterns and the higher at-

tempted correction (i.e., longer lasing times causing greater shifts in hydration). In further support of this explanation, the incidence of central islands, clinically, has been seen to decrease with increased suction of the laser plume, suggesting that corneal dehydration during the procedure mitigates against island formation. 19 Corneal hydration during the ablation also has been found to affect surface morphology after PRK.28 Differences in both epithelial and stromal wound healing resulting from differences both in the ablation contour as well as the configuration of the treated-nontreated corneal junction. 6 •7 •29 Patterns of epithelial migration and hyperplasia could account for early topography patterns (Fig 4).28 Both epithelial and stromal wound healing and remodeling, moreover, are likely important in temporal changes in the topography patterns seen. 29 - 33 In particular, the toric-with-axis and toric-against-axis patterns might indicate meridional differences in wound healing. 34 Wound healing also could explain the association that was found in this study between the irregular topography patterns and the higher attempted correction (i.e., greater corrections could be associated with relatively more exuberant early wound healing, thus leading to a more irregular topography pattern).

1340

Clinical and Optical Effects of 6 mrn versus Smaller Beam Diameters Although the two studies cannot be directly compared because the two patient cohorts were treated at different times and were not randomized, it is interesting to contrast the results found in our previous investigation of the 4.5to 5-mm treatment zone with the findings of the 6-mm zone described herein. In the study of corneal topography using the smaller ablation zone, there were significant associations of topography pattern with postoperative uncorrected visual acuity, predictability, refractive astigma-

Hersh et al . Corneal Topography of Excimer Laser PRK tism, and patient satisfaction. In general, the regular groups showed a better clinical outcome. 4 In that study, no association with topography pattern was found for best-spectacle-corrected visual acuity or glarelhalo. In comparison, the current study of topography using the 6-mm treatment zone showed no association with uncorrected visual acuity, spectacle-corrected visual acuity, astigmatism, corneal haze, glare, halo, or patient satisfaction. The lack of association with uncorrected visual acuity may be a result of the excellent mean uncorrected vision seen in this study in all of the topography patterns. The keyhole/semicircular group showed an average overcorrection that would not be expected if this pattern resulted from underablation of portions of the treatment zone as hypothesized in the hydration theory. Moreover, the keyhole/semicircular group was not associated statistically with glare or halo in this study. However, a small number of these patients did report substantial glare or halo. Although again not directly comparable statistically because of study design, the mean glarelhalo index in the 4.5- to 5-mm study was 1.89 (n = 159, standard deviation = 1.49, index = maximum of glare or halo on a scale of 0-5) compared with an average glarelhalo index of 1.33 (n = 91, standard deviation = 1.47) in this study of patients treated with a 6-mm beam diameter (P = 0.005, t test). In a prospective, randomized study comparing a 5mm treatment zone to a 6-mm zone, O'Brart et aC similarly found a significantly lower grade of halo in the 6mm group at I week, 1 month, and 6 months after 3.0 D corrections and at 1 week and 1 month in eyes with 6.0 D corrections. Such an improvement in subjective visual symptomatology also would be expected from theoretical optics?,3 Although there was no association of patient satisfaction with topography pattern, six patients (6.6%) did rank patient satisfaction at the lowest level of 0 or 1. Of these ranks, three patients ranked glare or halo at the highest level of 5. No obvious cause of dissatisfaction was found for the other three patients.

Conclusions Specific patterns of corneal topography appear to result from photorefractive keratectomy. 35 These patterns change with time, generally evolving toward an optically smoother corneal surface. 31 ,35 The diameter of the treatment zone, moreover, affects the topography pattern and may influence clinical outcomes. Although perturbative optical side effects seem diminished with the 6-mm treatment zone and most patients are satisfied with the procedure, a small number still report dissatisfaction with their result and report severe glare or halo. Operative and postoperative interventions to modulate corneal topography and further research to define the optimum corneal optical surface should help to improve objective and subjective visual outcomes of PRK in the future. Acknowledgments. The authors thank Joanne Katz, ScD, the Johns Hopkins University School of Public Health, for her assistance in data analysis and Deana Antonellis, Wini Dillon,

Maureen O'Connell, and Kimberley Doney, Summit Technology, Inc, for compiling clinical data.

Appendix The Summit PRK Topography Study Group List of Participants Binningham, AL: Marc Michelson, MD, John Owen, MD San Diego, CA: Michael Gordon, MD Boston, MA: Roger Steinert, MD, Carmen Puliafito, MD, Helen Wu, MD, Michael Raizman, MD St. Louis, MO: Jay Pepose, MD Kansas City, MO: Daniel Durrie, MD, Timothy Cavanaugh, MD, John Hunkeler, MD

References 1. Schwartz-Goldstein BH, Hersh PS. The Summit Photorefractive Keratectomy Topography Study Group. Corneal topography of phase III excimer laser photorefractive keratectomy: optical zone centration analysis. Ophthalmology 1995; 102:951-62. 2. Uozato H, Guyton DL. Centering corneal surgical procedures. Am J Ophthalmol 1987; 103:264-75. 3. Maloney RK. Corneal topography and optical zone location in photorefractive keratectomy. Refract Corneal Surg 1990;6:363-71. 4. Hersh PS, Schwartz-Goldstein BH. The Summit Photorefractive Keratectomy Topography Study Group. Corneal topography of phase III excimer laser photorefractive keratectomy: characterization and clinical effects. Ophthalmology 1995; 102:963-78. 5. Holladay JT, Prager TC. Mean visual acuity [letter]. Am J Ophthalmol 1991; 111:372-4. 6. Hersh PS, Shah SI, Geiger D, et al. Corneal optical irregularity after excimer laser photorefractive keratectomy. J Cataract Refract Surg 1996; 22: 197 - 204. 7. O'Brart DP, Corbett MC, Lohmann CP, et al. The effects of ablation diameter on the outcome of excimer laser photorefractive keratectomy. A prospective, randomized, doubleblind study. Arch Ophthalmol 1995; 113:438-43. 8. Shah SI, Hersh PS. Photorefractive keratectomy for myopia with a 6-mm beam diameter. J Refract Surg 1996; 12:3416.

9. Munnerlyn CR, Koons SJ, Marshall J. Photorefractive keratectomy: a technique for laser refractive surgery. J Cataract Refract Surg 1988; 14:46-52. 10. Blaker JW, Hersh PS. Theoretical and clinical effect of preoperative corneal curvature on excimer laser photorefractive keratectomy for myopia. J Refract Corneal Surg 1994; 10:571-4. 11. Hersh PS, Shah SI, Holladay JT, The Summit Photorefractive Keratectomy Topography Study Group. Corneal asphericity following excimer laser photorefractive keratectomy. Ophthalmic Surg Lasers 1996;27(5 Suppl):5421-8. 12. Seiler T, Genth U, Holschbach A, Derse M. Aspheric photorefractive keratectomy with excimer laser. Refract Corneal Surg 1993;9:166-72. 13. Maguire LJ, Bechara S. Epithelial distortions at the ablation zone margin after excimer laser photorefractive keratectomy from myopia. Am J Ophthalmol 1994; 117:809-10. 14. Hersh PS, Shah SI, Durrie D. Monocular diplopia following

1341

Ophthalmology

15.

16. 17. 18. 19. 20. 21. 22. 23.

24.

Volume 104, Number 8, August 1997

excimer laser photorefractive keratectomy after radial keratectomy. Ophthalmic Surg Lasers 1996;27:315-7. Roberts C. Characterization of the inherent error in a spherically-biased corneal topography system in mapping a radially aspheric surface. J Refract Corneal Surg 1994; 10: 1036. Roberts C. The accuracy of 'power' maps to display curvature data in corneal topography systems. Invest Ophthalmol Vis Sci 1994;35:3525-32. Mandell RE. The enigma of the corneal contour. CLAO J 1992; 18:267-73. Maguire LJ, Wilson SE, Camp JJ, Verity S. Evaluating the reproducibility of topography systems on spherical surfaces. Arch Ophthalmol 1993; 111:259-62. Mandell RB. Corneal power correction factor for photorefractive keratectomy. J Refract Corneal Surg 1994; 10: 1258. Klein SA, Mandell RB. Shape and refractive powers in corneal topography. Invest Ophthalmol Vis Sci 1995; 36:2096-109. Lin DT, Sutton HF, Berman M. Corneal topography following excimer photorefractive keratectomy for myopia. J Cataract Refract Surg 1993; 19(5uppl): 149-54. Levin S. Carson CA, Garrett SK, Taylor HR. Prevalence of central islands after excimer laser refractive surgery. J Cataract Refract Surg 1995;21:21-6. Colin J, Cochener B, Gallinaro C. Central steep islands immediately following excimer photorefractive keratectomy for myopia [letter]. Refract Corneal Surg 1993; 9:395-6. Puliafito CA, Stem D, Krueger RR, Mandel ER. Highspeed photography of excimer laser ablation of the cornea. Arch Ophthalmol 1987; 105:1255-9.

1342

25. Noack J, Tonnies R, Ohla H, et al. Influence of ablation plume dynamics on the formation of central islands of excimer laser PRK. Ophthalmology 1997; 104:823-30. 26. Krueger RR, Krasinski JS, Radzewicz C, et al. Photography of shock waves during excimer laser ablation of the cornea. Effect of helium gas on propagation velocity. Cornea 1993; 12:330-4. 27. Michalos P, Avila EN, Florakis GJ, Hersh PS. Do human tears absorb ultraviolet light? CLAO J 1994;20:192-3. 28. Simon G, Ren Q, Kervick GN, Parel JM. Optics of the corneal epithelium. Refract Corneal Surg 1993;9:42-50. 29. Wu WC, Stark WJ, Green WR. Corneal wound healing after 193-nm excimer laser keratectomy. Arch Ophthalmol 1991; 109:1426-32. 30. Carr J, Patel R, Hersh PS. Retreatment techniques for haze following photorefractive keratectomy. J Refractive Surg 1995; II(Suppl):309-13. 31. Wilson SE, Klyce SD, McDonald MB, et al. Changes in corneal topography after excimer laser photorefractive keratectomy for myopia. Ophthalmology 1991;98:1338-47. 32. Campos M, Trokel SL, McDonnell PJ. Surface morphology following photorefractive keratectomy. Ophthalmic Surg 1993; 24:822-5. 33. Grimm B, Waring GO III, Ibrahim O. Regional variation in corneal topography and wound healing following photorefractive keratectomy. J Refract Surg 1995; 11:348-57. 34. Shieh E, Moreira H, D' Arcy J, et al. Quantitative analysis of wound healing after cylindrical and spherical excimer laser ablations. Ophthalmology 1992; 99: 1050-5. 35. Klyce SD, Smolek MK. Corneal topography of excimer laser photorefractive keratectomy. J Cataract Refract Surg 1993; 19(5uppl):122-30.