- Email: [email protected]
Visual performance following photoastigmatic refractive keratectomy
Visual performance following photoastigmatic refractive keratectomy A prospective long-term study Weldon W. Haw, MD, Edward E. Manche, MD ABSTRACT Purpose: To prospectively determine the long-term effect of excimer laser photoastigmatic refractive keratectomy (PARK) on visual performance using psychophysical assessments and to identify predictors of poor performance on the psychophysical assessments. Setting: University-based hospital, Stanford University, Stanford, California, USA. Methods: Ninety-three eyes of 56 patients with a mean of ⫺4.98 diopters ⫾ 1.80 (SD) of primary compound myopic astigmatism had PARK using the Summit Apex Plus excimer laser and an erodible mask system. Patients were prospectively evaluated 1, 3, 6, 9, 12, 18, and 24 months following the procedure. Primary outcome variables included changes in the contrast sensitivity function curve (3.0, 6.0, 12.0, 18.0 cycles per degree) under 2 standard illuminance conditions (scotopic and photopic) and changes in the best spectacle-corrected visual performance under scotopic, photopic, and glare conditions. Results: A relative decline in the contrast sensitivity function curve occurred in the early postoperative period under both scotopic and photopic conditions. This was most pronounced under photopic illuminance and at the low spatial frequencies at the 6 month visit. By 1 year, however, the mean contrast sensitivity at all spatial frequencies and all illuminance conditions had returned to the preoperative level. Further improvements beyond the preoperative level may be related to the independent analysis of retreatment eyes beyond 6 months. A higher level of attempted correction of the spherical equivalent was predictive of an elevated scotopic contrast threshold at the extreme spatial frequencies 6 months after PARK (P ⬍ .05). The attempted level of astigmatic correction was predictive of a poor best corrected visual performance under scotopic conditions at 1 month (P ⬍ .05). This effect was only temporary and by postoperative month 3, there was no predictive effect of preoperative astigmatism (P ⬎ .05). Conclusions: Psychophysical assessments may be a more sensitive indicator of decreases in visual performance following excimer laser refractive surgery. The attempted level of correction of spherical equivalent and astigmatism may adversely affect early scotopic visual performance. Decreases in visual performance are temporary, return to normal by 12 months, and remain stable 24 months following PARK. J Cataract Refract Surg 2000; 26:1463–1472 © 2000 ASCRS and ESCRS
© 2000 ASCRS and ESCRS Published by Elsevier Science Inc.
0886-3350/00/$–see front matter PII S0886-3350(00)00309-6
VISUAL PERFORMANCE AFTER PARK
M
ost studies emphasize the use of uncorrected visual acuity (UCVA) and refractive change in measuring the success of refractive surgery procedures. It has been suggested that high-contrast Snellen visual acuity may be inappropriate and too insensitive a barometer to measure the subtle decrements in visual function after photorefractive keratectomy (PRK).1–3 Thus, many investigations have been directed at refining more sensitive assessments of visual performance following refractive surgery.1–16 These studies have investigated PRK’s effect on psychophysical assessments such as contrast sensitivity,1,2,4,9 –11,13–16 low-contrast visual acuity,3,5,12 near (reading) contrast sensitivity,8 “night vision,”6 –7 and disability glare.1,2,9,10 –12,14 These tests have been used on patients who have had different refractive surgical procedures including PRK and radial keratotomy (RK). However, more recent technology has permitted the use of the excimer laser to correct myopic astigmatism. In photoastigmatic refractive keratectomy (PARK) or toric PRK, a slit or elliptical aperture can be used to produce an astigmatic correction. When used in combination with erodible mask technology, an effective reduction in compound myopic astigmatism may be achieved.17 Ablation through an erodible mask may also produce a smoother and clearer cornea than that achieved with the diaphragm.18 A smoother and clearer cornea may be expected to produce few aberrations and have a smaller impact on contrast sensitivity declines. In this prospective long-term study, we evaluated the visual performance following PARK with the Summit Apex Plus excimer laser administered through an erodible mask system. In addition, we attempted to identify predictors/prognostic factors for poor visual performance on these psychophysical tests.
Patients and Methods As part of a U.S. Food and Drug Administration phase III clinical trial, 93 eyes of 56 patients with priAccepted for publication June 19, 2000. From Stanford University School of Medicine, Stanford, California, USA. Neither author has a financial interest in any product mentioned. Reprint requests to Edward E. Manche, MD, Stanford University School of Medicine, 300 Pasteur Drive, Suite A175, Stanford, California 94305, USA. 1464
mary compound myopic astigmatism were enrolled at a single center (Stanford University). Inclusion criteria were a myopic spherical refraction of ⫺1.0 to ⫺7.0 diopters (D), myopic cylindrical refraction of ⫺1.0 to ⫺5.0 D, age of at least 18 years, preoperative best spectacle-corrected visual acuity (BSCVA) of 20/25 or better, and stable refractive error within ⫾1.0 D for 1 year. Patients were excluded from the study if they had clinically significant ocular pathology and underlying systemic diseases known to affect corneal health; were functionally monocular or taking medications adversely affecting wound healing (ie, steroids); had participated in prior ophthalmic clinical trials; had a difference of ⬎1.0 D between preoperative manifest refraction and cycloplegic refraction, intraocular pressure of ⬎21.0 mm Hg, or previous ocular surgery. Patients with corneal warpage secondary to contact lens wear were identified by preoperative corneal topography and were excluded until their serial topography returned to normal. All patients completed an informed consent form approved by the Institutional Review Board at Stanford University after an explanation of the nature and possible consequences of the study. The preoperative examination included a complete ocular history and examination including refraction, slitlamp biomicroscopy, corneal topography, and contrast sensitivity measurements under 2 standard ambient illuminance conditions (21 and 324 lux) and 4 spatial frequencies (3.0, 6.0, 12.0, and 18.0 cycles per degree [cpd]). Measurements of BSCVA under photopic (324 lux), scotopic (21 lux), and glare conditions were also performed using a Lighthouse Distance Visual Acuity Test, second edition (Lighthouse Low Vision Products). This uses a chart modified from the Early Treatment of Diabetic Retinopathy Study with Sloan letters. Glare conditions were simulated by a handheld brightness acuity test meter (Mentor) at its maximum setting. Visual acuity is reported as Snellen equivalents. The term best spectacle-corrected visual performance will be used to refer to these measurements. Topical proparacaine hydrochloride 0.5% (Ophthetic威), ofloxacin 0.3% (Ocuflox威), and diclofenac sodium 0.1% (Voltaren威) were administered immediately prior to the procedure. Appropriate alignment was achieved by marking the horizontal axis (0 to 180 degrees) with a surgical marking pen (Devon Industries), using a horizontal slit beam at the slitlamp with patients
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
VISUAL PERFORMANCE AFTER PARK
spatial frequencies (3.0, 6.0, 12.0, and 18.0 cpd). Contrast sensitivity measurements were made with the best spectacle correction in place. Each patient completed a survey rating adverse symptoms such as glare, halos, visual blurring, and color distortion. Each individual symptom was scored on a scale of 0 (none) to 5 (severe). This survey was conducted during the preoperative examination and at the 6, 12, and 24 month postoperative visits. Additionally, corneal clarity was assessed by slitlamp biomicroscopy and recorded on a scale of 0 to 5 by a single observer (E.E.M.). The SPSS Graduate Pack for Windows V6.1.3 (SPSS Inc.) was used to determine statistical significance. The analysis of predictors of poor visual performance was performed using a regression model. Differences were considered to be statistically significant when the P value was ⬍.05.
seated in the upright position. Patients were taken to the laser room and placed beneath the laser in the supine position. A wire speculum was placed in the eye being treated. The corneal epithelium was manually removed with a blunt PRK spatula. All patients had PARK at Stanford University Medical Center using the 193 nm Summit Apex Plus excimer laser delivered through a poly(methyl methacrylate) erodible mask system. Standard laser parameters were used in all cases. These included a repetition rate of 10 Hz, radiant exposure at the corneal plane of 180 mJ/cm2, and a 6.5 ⫻ 5.0 mm elliptical ablation zone. The ablation was delivered through an in-the-rail erodible mask provided by Summit Technology. Attempted correction was based on the manifest refraction. Postoperatively, a bandage contact lens was placed on the eye until the cornea had completely re-epithelialized. Postoperative eye medications included topical Ocuflox 4 times a day for 4 days or until the epithelium was healed and topical fluorometholone 0.1% (FML威) 4 times a day for 1 month and then 2 times a day for 1 month. Eyes were prospectively assessed at 1, 3, 6, 9, 12, 18, and 24 months following the procedure. All postoperative visits included a measurement of BSCVA under 3 standard conditions: scotopic (approximately 21 lux), photopic (approximately 324 lux), and glare. Contrast sensitivity measurements were performed 6, 12, and 24 months after the PARK procedure. Contrast sensitivity was evaluated using the CSV1000E (Vector Vision). The CSV-1000E uses a chart with vertical sine wave grating patterns. Each pattern represents an increase in contrast sensitivity determined by the manufacturer. The contrast sensitivity is recorded in log units. There are 9 contrast levels for each of the 4
Results The mean patient age was 41.4 years ⫾ 8.54 (SD) (range 26 to 58 years). Forty-nine eyes (52.7%) were in women and 44 (47.3%), in men. Prior to the 2 year follow-up, 24 eyes (25.8%) required retreatment. Seven eyes had LASIK enhancements and 17, PRK/PARK retreatment. These eyes were evaluated independently from the 2 year follow-up results. At the 2 year followup, 59 of the remaining eyes (71.1%) were available for analysis. Refractive results are shown in Table 1 and UCVA, in Table 2. Refractive and UCVA results for the subset of nonretreated eyes available at 1, 3, 6, 12, 18, and 24 months are shown in Table 3. All other visual acuity results are reported as best spectacle-corrected visual performance unless otherwise specified.
Table 1. Mean (⫾SD) sphere, cylinder, and SE at each of the designated intervals. Months Postop Preop
1
3
6
9
12
18
Sphere (D)
⫺3.88 ⫾ 1.83
⫺0.20 ⫾ 0.69
⫺0.33 ⫾ 0.75
⫺0.23 ⫾ 0.88
⫺0.22 ⫾ 0.88
⫺0.26 ⫾ 0.82
⫺0.16 ⫾ 0.92
0.05 ⫾ 0.86
Cylinder (D)
⫺2.19 ⫾ 0.90
⫺0.37 ⫾ 0.42
⫺0.41 ⫾ 0.45
⫺0.61 ⫾ 0.48
⫺0.64 ⫾ 0.52
⫺0.76 ⫾ 0.52
⫺0.75 ⫾ 0.52
⫺0.89 ⫾ 0.58
SE (D)
⫺4.98 ⫾ 1.80
⫺0.38 ⫾ 0.67
⫺0.53 ⫾ 0.68
⫺0.75 ⫾ 0.85
⫺0.54 ⫾ 0.81
⫺0.65 ⫾ 0.73
⫺0.53 ⫾ 0.80
⫺0.39 ⫾ 0.72
93
92
88
85
66
72
52
59
Number of eyes
24
SD ⫽ standard deviation; SE ⫽ spherical equivalent
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
1465
VISUAL PERFORMANCE AFTER PARK
Table 2. Uncorrected visual acuity results (number of eyes [%]) at each postoperative interval. Months Postoperative UCVA
1
3
6
9
12
18
24
ⱕ20/20
74 (80)
51 (58)
48 (56)
35 (53)
36 (50)
28 (54)
34 (58)
ⱕ20/40
87 (94)
78 (89)
70 (82)
57 (86)
67 (93)
48 (92)
56 (95)
Table 3. Refraction and UCVA results in subset of eyes not treated and available at specified months. Months Postop Preop
1
3
6
9
18
24
Sphere (D)
⫺3.78 ⫾ 1.90
⫺10.00 ⫾ 0.62
⫺0.14 ⫾ 0.64
⫺0.10 ⫾ 0.61
⫺0.71 ⫾ 0.77
⫺0.07 ⫾ 0.88
0.01 ⫾ 0.86
Cylinder (D)
⫺2.26 ⫾ 0.90
⫺0.33 ⫾ 0.43
⫺0.43 ⫾ 0.49
⫺0.58 ⫾ 0.47
⫺0.69 ⫾ 0.49
⫺0.78 ⫾ 0.51
⫺0.81 ⫾ 0.51
SE (D)
⫺4.94 ⫾ 1.91
⫺0.27 ⫾ 0.56
⫺0.36 ⫾ 0.48
⫺0.46 ⫾ 0.56
⫺0.52 ⫾ 0.65
⫺0.46 ⫾ 0.77
⫺0.39 ⫾ 0.74
ⱕ20/20
30 (60)
32 (64)
32 (64)
35 (70)
31 (62)
30 (60)
30 (60)
ⱕ20/40
48 (96)
44 (88)
45 (90)
46 (92)
50 (100)
47 (94)
47 (94)
Sphere, cylinder, and SE results are mean ⫾ SD; visual acuity results are number of eyes (%).
Contrast Sensitivity Function The spatial frequency and standard illuminance condition (scotopic and photopic) are specified for each figure. Tables 4 and 5 show the change in contrast sensitivity under both illuminance conditions and all measured spatial frequencies. Six months after the PARK treatment, the mean contrast sensitivity at all spatial frequencies under the photopic contrast measurements and at all low spatial frequencies (3.0 and 6.0 cpd) under the scotopic contrast measurements remained the same or decreased from the preoperative level. The decrease was statistically significant (P ⬍ .05) at the low spatial frequen-
cies under the photopic condition and at the lowest spatial frequency (3.0 cpd) under the scotopic condition. By 1 year, however, the mean contrast sensitivity at all spatial frequencies and all illuminance conditions had returned to the preoperative level (Tables 4 and 5). At all spatial frequencies and both illuminance conditions, the 12 and 24 month mean contrast sensitivities were higher than the 6 month mean contrast sensitivities. The mean contrast sensitivity at the 24 month visit was higher than the preoperative level under all conditions except the scotopic 3.0 cpd condition, which showed a statistically insignificant decrease in the mean contrast sensitivity of ⫺0.01 log units.
Table 4. Change in scotopic contrast sensitivity at each spatial
Table 5. Change in photopic contrast sensitivity at each spatial
frequency.
frequency.
Spatial Frequency (cpd)
24
Spatial Frequency (cpd)
6
12
6
12
24
3.0
⫺0.05 ⫾ 0.20
⫹0.02 ⫾ 0.26
⫺0.01 ⫾ 0.22
3.0
⫺0.07 ⫾ 0.19
⫹0.02 ⫾ 0.26
⫹0.01 ⫾ 0.21
6.0
0.00 ⫾ 0.31
⫹0.04 ⫾ 0.25
⫹0.08 ⫾ 0.23
6.0
⫺0.14 ⫾ 0.35
⫹0.04 ⫾ 0.25
⫹0.06 ⫾ 0.41
12.0
⫹0.18 ⫾ 0.65
⫹0.16 ⫾ 0.50
⫹0.21 ⫾ 0.55
18.0
⫹0.03 ⫾ 0.40
⫹0.20 ⫾ 0.40
⫹0.23 ⫾ 0.43
Months Postoperative
(P ⫽ .03)
(P ⬍ .01)
Change in contrast sensitivity (log units) is calculated from the preoperative contrast sensitivity measurement.
1466
Months Postoperative
(P ⬍ .01) 12.0
⫺0.04 ⫾ 0.51
⫹0.16 ⫾ 0.50
⫹0.09 ⫾ 0.44
18.0
0.00 ⫾ 0.40
⫹0.20 ⫾ 0.40
⫹0.13 ⫾ 0.35
Change in contrast sensitivity (log units) is calculated from the preoperative contrast sensivity measurement.
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
VISUAL PERFORMANCE AFTER PARK
Figure 1. (Haw) Mean change in BSCVA under scotopic conditions. Error bars represent 95% confidence interval.
Figure 2. (Haw) Mean change in BSCVA under photopic condi-
Figure 3. (Haw) Mean change in BSCVA under glare conditions. Error bars represent 95% confidence interval.
Figure 4. (Haw) Loss of BSCVA under scotopic conditions at
tions. Error bars represent 95% confidence interval.
24 months.
Figure 5. (Haw) Loss of BSCVA under photopic conditions at
Figure 6. (Haw) Loss of BSCVA under glare conditions at
24 months.
24 months.
Best Spectacle-Corrected Visual Performance Figures 1 to 3 show the mean change (lines reported in Snellen equivalents) in the best spectaclecorrected visual performance throughout the postoperative period under scotopic, photopic, and glare conditions. There was a mean decrease of ⫺0.56 ⫾ 1.19 lines 1 month after PARK under the scotopic condition (Figure 1). The scotopic best spectacle-corrected visual performance increased at every measured postoperative period. By 6 months, the best corrected visual performance reached the preoperative level. Figure 2 demonstrates a similar recovery in the photopic best corrected visual performance. At 1 month, there was a mean decrease of 0.60 ⫾ 1.10 lines. This also improved at every measured postoperative interval. The
photopic best corrected visual performance returned to the preoperative level by 9 months. As in the scotopic best corrected visual performance, recovery was more rapid between 1 and 3 months. Figure 3 demonstrates the mean change in the glare best corrected visual performance. At 1 month, there was a mean decline of ⫺0.33 ⫾ 1.32 lines. The glare best corrected visual performance returned to the preoperative level between 3 and 6 months. As in the scotopic and photopic best corrected visual performance conditions, there was an improvement at every measured postoperative visit. At the last follow-up, 0 eyes lost ⬎2 lines of best corrected visual performance under photopic or glare conditions, while 1 eye (1.7%) lost ⬎2 lines under scotopic conditions (Figures 4 to 6).
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
1467
VISUAL PERFORMANCE AFTER PARK
Visual Performance After Retreatment Twenty-four eyes (25.8%) had retreatment for residual refractive error. Insufficient results are available on the 7 eyes that had LASIK retreatment. Therefore, analysis is confined to the 17 eyes of 14 patients who had retreatment with PRK or PARK. The mean retreatment was 11.3 ⫾ 4.9 months (range 6 to 23 months) following the original PARK treatment. Eight eyes (47%) had PARK enhancement for residual compound myopic astigmatism; 9 eyes (53%) had PRK enhancement for residual spherical myopia. Prior to enhancement, the mean preoperative sphere was ⫺1.62 ⫾ 0.75 D (range ⫺0.75 to ⫺3.25 D), the mean preoperative astigmatism was ⫺0.93 ⫾ 0.64 D (range 0.00 to ⫺2.50 D), and the mean spherical equivalent was ⫺2.08 ⫾ 0.76 D (range ⫺1.13 to ⫺3.75 D). At the last follow-up (mean 12.9 months), the mean sphere was ⫺0.29 ⫾ 1.23 D, the mean astigmatism was ⫺0.88 ⫾ 0.40 D, and the mean spherical equivalent was – 0.74 ⫾ 1.27 D. The UCVA was 20/20 or better in 59% of eyes and 20/40 or better in 88%. There was no significant (P ⬎ .05) decrease in the mean contrast sensitivity from the pre-retreatment levels at all spatial frequencies and both illuminance conditions 6 months after retreatment. Similarly, no significant decrease was noted at the last follow-up. The mean contrast sensitivity under both illuminance conditions was not significantly lower than the mean preoperative levels. Four eyes (24%) experienced haze at the last postoperative visit following retreatment; 2 of them developed late-onset haze and regression. Among the 4 eyes, 2 lost 2 lines of best corrected visual performance under glare conditions, and 1 lost 2 lines of best corrected visual performance under both scotopic and photopic conditions. The worst BSCVA was 20/32 under any of these conditions. No eye lost more than
2 lines of best corrected visual acuity under scotopic, photopic, or glare conditions. Corneal Haze Corneal haze was graded on a scale of 0 to 5 at the slitlamp. The mean level of corneal haze increased progressively up to the 12 month follow-up. The corneal haze diminished at the 18 and 24 month follow-ups. The mean corneal haze score decreased 47% from its peak value of 0.32 ⫾ 0.80 at 12 months to 0.17 ⫾ 0.46 at 24 months. Eight eyes (13.5%) had trace to mild haze and no eye had moderate or severe haze at the 24 month follow-up. At the point of maximum corneal haze (12 months), 11 eyes (15.3%) demonstrated grade 1 to grade 3 haze. These eyes demonstrated a loss of best spectacle-corrected visual performance of ⫺0.64 ⫾ 1.50 lines under scotopic conditions, ⫺0.27 ⫾ 0.90 lines under photopic conditions, and ⫺0.09 ⫾ 0.94 lines under glare conditions. This compared adversely with the mean increase in best spectacle-corrected visual performance in eyes without haze in which the respective values were ⫹0.49 ⫾ 1.09 (P ⫽ .09), ⫹0.26 ⫾ 1.21 (P ⫽ .62), and ⫹0.82 ⫾ 1.16 (P ⫽ .24). At 24 months, the mean best corrected visual performances under scotopic and glare conditions remained lower in eyes with haze than in those without haze (Table 6). However, these values were not statistically significant. Eyes with haze had a tendency to demonstrate a more pronounced decrease in the contrast sensitivity function curve under lower spatial frequencies (3.0 and 6.0 cpd) than eyes with no corneal haze at the 6 month visit. However, this was only statistically significant for the 6.0 cpd data point under photopic conditions (P ⫽ .02). By 12 months, there was no statistically significant difference between eyes with or without haze in the contrast sensitivity function curve at all measured spatial
Table 6. Change in best corrected visual performance in eyes with and without haze under each condition. 6 Months Condition Scotopic Photopic Glare
12 Months
24 Months
ⴙ Haze
ⴚ Haze
ⴙ Haze
ⴚ Haze
ⴙ Haze
ⴚ Haze
⫺0.13 ⫾ 0.74
⫹0.10 ⫾ 1.04
⫺0.65 ⫾ 1.50
⫹0.49 ⫾ 1.09
⫹0.25 ⫾ 0.89
⫹0.64 ⫾ 0.94
0.00 ⫾ 1.00
⫺0.29 ⫾ 1.03
⫺0.27 ⫾ 0.90
⫺0.27 ⫾ 0.90
⫹0.50 ⫾ 0.76
⫹0.48 ⫾ 0.84
⫹0.19 ⫾ 0.98
⫹0.33 ⫾ 1.24
⫺0.09 ⫾ 0.94
⫹0.82 ⫾ 1.16
⫹0.25 ⫾ 0.85
⫹0.78 ⫾ 0.99
Mean change in best corrected visual performance recorded in change of lines (Snellen equivalents) from the preoperative value.
1468
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
VISUAL PERFORMANCE AFTER PARK
Figure 7. (Haw) Change in contrast sensitivity in eyes with and without haze under scotopic and photopic conditions.
frequencies and illuminance conditions. At the last follow-up at 24 months, there was a slight tendency toward a lower contrast sensitivity in eyes with haze than in eyes without haze at all but the highest spatial frequency (18.0 cpd) (Figure 7). At this postoperative interval, only the photopic 3.0 cpd (P ⬍ .01) data point demonstrated a statistically significant decrease between eyes with and without haze. Survey of Symptoms The increases in the mean glare, visual blurring, and color distortion scores were not statistically significant (P ⬎ .05) at the last postoperative visit. Despite an improvement in the halo score from months 18 to 24, there was a statistically significant increase in the halos from the preoperative level. The preoperative score of 0.64 ⫾ 0.98 increased to 1.09 ⫾ 1.20 at the 24 month follow up (P ⫽ .04). Predictors of Poor Visual Performance A regression model was used to evaluate potential predictors for the relative decline in the visual performance within the first 6 months, prior to the exclusion of retreatment eyes. A regression was also used to determine whether a change in the glare score, halo score, visual blurring score, and color distortion score predicted declines in the visual performance over the same interval. The effect of the level of spherical equivalent correction and the level of astigmatic correction on the contrast sensitivity function curve was specifically evaluated. The photopic contrast sensitivity function curve was not significantly affected (P ⬎ .05) by either preoperative variable. However, eyes with high preoperative spherical equivalents were predictive of an elevated scotopic con-
trast threshold at the extreme spatial frequencies (3.0 cpd [P ⫽ .02] and 18.0 cpd [P ⫽ .03]) 6 months after PARK. The best corrected visual performance under glare and photopic conditions at 1, 3, and 6 months was not affected by the level of preoperative spherical equivalent or the level of preoperative astigmatism (P ⬎ .05). However, poor best corrected visual performance under scotopic conditions at 1 month was predicted by the level of preoperative astigmatism (P ⬍ .05). This effect was only temporary and by postoperative month 3, there was no predictive effect of preoperative astigmatism (P ⬎ .05). The changes in the glare, halo, visual blurring, and color distortion scores did not predict changes in either the contrast sensitivity function curve under scotopic and photopic conditions or the best corrected visual performance under glare, scotopic, and photopic conditions (P ⬎ .05). Vector analysis performed at the 24 month follow-up demonstrated no predictive effect of surgically induced astigmatism, target induced astigmatism, mean angle of error, difference vector, or mean magnitude of error on the loss of BSCVA under glare, photopic, or scotopic conditions. A higher mean angle of error was weakly predictive of lower contrast sensitivity under scotopic illuminance at 12.0 cpd (P ⫽ .047). No other vector analysis variable was predictive of poor contrast sensitivity (P ⬎ .05).
Discussion There is substantial evidence that contrast sensitivity is adversely affected following refractive surgical procedures. Some studies demonstrate a measurable decrease in the contrast sensitivity that returns to normal 6 months to 1 year after the refractive surgical procedure.8,14,16 However, other studies suggest that visual function on psychophysical tests may still be reduced at or beyond 1 year.2–3,9,12 In our cohort, there was a tendency for a relative decline in the contrast sensitivity function curve in the early postoperative period under both scotopic and photopic conditions. This was most pronounced under photopic illuminance conditions, as well as at the low spatial frequencies at the 6 month visit. The recovery of mean contrast sensitivity function was most profound at the higher spatial frequencies and was demonstrated by a
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
1469
VISUAL PERFORMANCE AFTER PARK
tendency toward improvement in contrast sensitivity at the 6 month visit and by a statistically significant increase in contrast sensitivity at 12 month. By 12 months, the mean contrast sensitivity had returned to normal at all spatial frequencies and under both illuminance conditions. The contrast sensitivity remained stable at the 24 month visit. We noted a tendency toward decreased performance on the contrast sensitivity function curve under photopic conditions at the 6 month follow-up. Although the physiologic pupillary dilation that occurs during scotopic illumination may allow optical irregularities outside the photopic optical zone to affect the contrast perception, this may have been offset by the background photopic illuminance, which may inherently decrease the contrast of the black letters on the back-lighted contrast chart. However, the best spectaclecorrected visual performance under scotopic and photopic conditions was less sensitive to this difference. Figures 1 and 2 mirrored each other during the postoperative time course. The general improvement in contrast from preoperatively may not be due to a “real” increase in contrast sensitivity following PARK but rather a “learning” effect. This phenomenon is well documented in patients taking difficult examinations such as automated visual field tests in which an artifactual deficit may improve on a repeat examination. This may also occur following the measurement of contrast sensitivity. Patients may “learn” to improve their performance on subsequent tests by repeat examinations. Since the patients had not had previous trials at measuring the contrast sensitivity, the preoperative level may be expected to be the lowest. Nevertheless, it is difficult to determine to what degree (if any) this phenomenon accounted for our results. Certainly, age-matched control groups that had not had PARK but had had repeat contrast sensitivity testing would have been useful in determining the degree to which this effect accounted for our observed results. In addition, we introduced a selection bias by independently analyzing a significant number of eyes that were retreated (26%). We do not know whether this selection bias may have contributed to the observed trends in visual function since the retreatment eyes did not experience a statistically significant decrease in their best spectacle-corrected contrast sensitivity function. 1470
The best corrected visual performance under scotopic, photopic, and glare conditions significantly decreased 1 month after PARK. However, there was a steady and progressive improvement in the best corrected visual performance under all 3 conditions at every measurable postoperative visit from 1 to 24 months. The recovery was most rapid in the early postoperative period between 1 and 3 months. Glare recovery was more rapid, with full recovery occurring at approximately 3 months rather than 6 to 9 months with the photopic and scotopic best corrected visual performances. The mean improvement in scotopic, photopic, and glare best corrected visual performances remained stable through the 24 month follow-up. Our regression analysis suggests that the level of preoperative spherical equivalent and the level of preoperative astigmatism were not predictive of a poor outcome on the photopic contrast sensitivity function curve 6 months after PARK. Although the level of preoperative astigmatism did not adversely affect the scotopic contrast sensitivity function curve, it was affected by the degree of preoperative spherical equivalent at the extreme spatial frequencies (3.0 and 18.0 cpd) during the same interval. This is not entirely surprising as a higher preoperative spherical equivalent corresponds to a larger ablation depth and a steeper transition between the unablated and ablated zone (6.5 ⫻ 5.0 mm ablation zone in our trial). This steeper transition would be expected to play a larger role under scotopic (physiologic pupillary dilation) conditions. It has been suggested that the Vector Vision CSV-1000 system used to measure contrast sensitivity in our study may unmask aberrations from the transition zone of ablated and unablated corneas in PRK.15 Similarly, the attempted level of astigmatic correction was predictive of an early poor outcome in the scotopic best corrected visual performance at 1 month. However, this had resolved by the 3 month postoperative visit. A more complicated astigmatic ablation may result in larger contrast sensitivity declines than a spherical ablation. However, this may be offset by the use of erodible mask technology, which results in a smoother ablation profile that would be expected to diminish the effect of optical aberrations resulting from an irregular ablation. In a study of 30 eyes that had RK and 30 eyes that had PRK, contrast sensitivity and glare disability testing
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
VISUAL PERFORMANCE AFTER PARK
did not accurately reflect patients’ subjective assessment of their visual performance in daily life.14 Our study also did not find the degree of subjective symptoms of glare, halo, blurring, or color distortion to accurately predict the declines in contrast sensitivity function or best corrected visual performance under scotopic, photopic, or glare conditions. Verdon and coauthors3 reported 18 patients who had PRK for –5.08 ⫾ 1.63 D of myopia. The authors concluded that there was no notable relationship between corneal haze and any visual performance measure. However, they cautioned that this might have resulted from the coarse grading scale used to assess haze, a restricted range of haze values, and a homogeneous and small sample size. In our study, eyes with corneal haze demonstrated a trend toward decreased contrast sensitivity at the lower spatial frequencies under both scotopic and photopic illuminance conditions at the earliest (6 month) measurement. Despite the tendency for improvement in the mean contrast sensitivity of treated eyes from the 6 month interval, eyes with persistent haze at the 24 month interval retained a lower mean contrast sensitivity at the lower (3.0 to 12.0 cpd) spatial frequencies. These eyes also demonstrated a lower best corrected visual performance than eyes without haze under scotopic and glare conditions by the 24 month followup. However, the photopic best corrected visual performance had returned to normal by this time. Retreatment with PRK or PARK did not adversely affect contrast sensitivity under scotopic or photopic illuminance conditions at any time during the postretreatment period. When compared with the mean preoperative contrast sensitivity prior to the original PARK procedure, there was also no statistically significant decrease in the contrast sensitivity under both illuminance conditions. However, best spectacle-corrected visual performance was adversely effected by late-onset haze in 2 retreated eyes. Our results appear to suggest a measurable decline in many of our psychophysical measurements with early, rapid improvements and return to the preoperative state by 6 months to 1 year following the treatment of compound myopic astigmatism with PARK. Possible predictors of early poor visual performance under scotopic conditions include attempted level of correction for spherical equivalent and astigmatism. Corneal haze may also adversely affect the best spectacle-corrected visual
performance under scotopic and glare conditions. We found a poor correlation between the patient’s subjective grading of symptoms with the patient’s objective performance on our psychophysical measurements. The psychophysical assessments tended to remain stable up to 24 postoperative months.
References 1. Ambrosio G, Cennamo G, De Marco R, et al. Visual function before and after phtorefractive keratectomy for myopia. J Refract Corneal Surg 1994; 10:129 –136 2. Butuner Z, Elliott DB, Gimbel HV, Slimmon S. Visual function one year after excimer laser photorefractive keratectomy. Refract Corneal Surg 1994; 10:625– 630 3. Verdon W, Bullimore M, Maloney R. Visual performance after photorefractive keratectomy, a prospective study. Arch Ophthalmol 1996; 114:1465–1472 4. Dutt S, Steinert RF, Raizman MB, Puliafito CA. Oneyear results of excimer laser photorefractive keratectomy for low to moderate myopia. Arch Ophthalmol 1994; 112:1427–1436 5. Lohmann CP, Gartry DS, Muir MK, et al. Corneal haze after excimer laser refractive surgery: objective measurements and functional implications. Eur J Ophthalmol 1991; 1:173–180 6. O’Brart DPS, Lohmann CP, Fitzke FW, et al. Disturbances in night vision after excimer laser photorefractive keratectomy. Eye 1994; 8:46 –51 7. O’Brart DPS, Lohmann CP, Fitzke FW, et al. Night vision after excimer laser photorefractive keratectomy: haze and halos. Eur J Ophthalmol 1994; 4:43–51 8. Hodkin MJ, Lemos MM, McDonald MB, et al. Near vision contrast sensitivity after photorefractive keratectomy. J Cataract Refract Surg 1997; 23:192–195 9. Niesen U, Businger U, Hartmann P, et al. Glare sensitivity and visual acuity after excimer laser photorefractive keratectomy for myopia. Br J Ophthalmol 1997; 81: 136 –140 10. Corbett MC, Prydal JI, Verma S, et al. An in vivo investigation of the structures responsible for corneal haze after photorefractive keratectomy and their effect on visual function. Ophthalmology 1996; 103:1366 –1380 11. Deitz MR, Piebenga LW, Matta CS, et al. Ablation zone centration after photorefractive keratectomy and its effect on visual outcome. J Cataract Refract Surg 1996; 22: 696 –701 12. Niesen UM, Businger U, Schipper I. Disability glare after excimer laser photorefractive keratectomy for myopia. J Refract Surg 1996; 12:S267–S268 13. Ficker LA, Bates AK, Steele AD McG, et al. Excimer laser photorefractive keratectomy for myopia: 12 month follow-up. Eye 1993; 7:617– 624 14. Ghaith AA, Daniel J, Stulting RD, et al. Contrast sensi-
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
1471
VISUAL PERFORMANCE AFTER PARK
tivity and glare disability after radial keratotomy and photorefractive keratectomy. Arch Ophthalmol 1998; 116: 12–18 15. Wachler BSB, Durrie DS, Assil KK, Krueger RR. Role of clearance and treatment zones in contrast sensitivity: significance in refractive surgery. J Cataract Refract Surg 1999; 25:16 –23 16. Wang Z, Chen J, Yang B. Comparison of laser in situ keratomileusis and photorefractive keratectomy to cor-
1472
rect myopia from ⫺1.25 to ⫺6.00 diopters. J Refract Surg 1997; 13:528 –534 17. Carones F, Brancato R, Morico A, et al. Compound myopic astigmatism correction using a mask in-the-rail excimer laser delivery system. Preliminary results. Eur J Ophthalmol 1996; 6:221–233 18. Maloney RK, Friedman M, Harmon T, et al. A prototype erodible mask delivery system for the excimer laser. Ophthalmology 1993; 100:542–549
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
© 2000 ASCRS and ESCRS Published by Elsevier Science Inc.
0886-3350/00/$–see front matter PII S0886-3350(00)00309-6
VISUAL PERFORMANCE AFTER PARK
M
ost studies emphasize the use of uncorrected visual acuity (UCVA) and refractive change in measuring the success of refractive surgery procedures. It has been suggested that high-contrast Snellen visual acuity may be inappropriate and too insensitive a barometer to measure the subtle decrements in visual function after photorefractive keratectomy (PRK).1–3 Thus, many investigations have been directed at refining more sensitive assessments of visual performance following refractive surgery.1–16 These studies have investigated PRK’s effect on psychophysical assessments such as contrast sensitivity,1,2,4,9 –11,13–16 low-contrast visual acuity,3,5,12 near (reading) contrast sensitivity,8 “night vision,”6 –7 and disability glare.1,2,9,10 –12,14 These tests have been used on patients who have had different refractive surgical procedures including PRK and radial keratotomy (RK). However, more recent technology has permitted the use of the excimer laser to correct myopic astigmatism. In photoastigmatic refractive keratectomy (PARK) or toric PRK, a slit or elliptical aperture can be used to produce an astigmatic correction. When used in combination with erodible mask technology, an effective reduction in compound myopic astigmatism may be achieved.17 Ablation through an erodible mask may also produce a smoother and clearer cornea than that achieved with the diaphragm.18 A smoother and clearer cornea may be expected to produce few aberrations and have a smaller impact on contrast sensitivity declines. In this prospective long-term study, we evaluated the visual performance following PARK with the Summit Apex Plus excimer laser administered through an erodible mask system. In addition, we attempted to identify predictors/prognostic factors for poor visual performance on these psychophysical tests.
Patients and Methods As part of a U.S. Food and Drug Administration phase III clinical trial, 93 eyes of 56 patients with priAccepted for publication June 19, 2000. From Stanford University School of Medicine, Stanford, California, USA. Neither author has a financial interest in any product mentioned. Reprint requests to Edward E. Manche, MD, Stanford University School of Medicine, 300 Pasteur Drive, Suite A175, Stanford, California 94305, USA. 1464
mary compound myopic astigmatism were enrolled at a single center (Stanford University). Inclusion criteria were a myopic spherical refraction of ⫺1.0 to ⫺7.0 diopters (D), myopic cylindrical refraction of ⫺1.0 to ⫺5.0 D, age of at least 18 years, preoperative best spectacle-corrected visual acuity (BSCVA) of 20/25 or better, and stable refractive error within ⫾1.0 D for 1 year. Patients were excluded from the study if they had clinically significant ocular pathology and underlying systemic diseases known to affect corneal health; were functionally monocular or taking medications adversely affecting wound healing (ie, steroids); had participated in prior ophthalmic clinical trials; had a difference of ⬎1.0 D between preoperative manifest refraction and cycloplegic refraction, intraocular pressure of ⬎21.0 mm Hg, or previous ocular surgery. Patients with corneal warpage secondary to contact lens wear were identified by preoperative corneal topography and were excluded until their serial topography returned to normal. All patients completed an informed consent form approved by the Institutional Review Board at Stanford University after an explanation of the nature and possible consequences of the study. The preoperative examination included a complete ocular history and examination including refraction, slitlamp biomicroscopy, corneal topography, and contrast sensitivity measurements under 2 standard ambient illuminance conditions (21 and 324 lux) and 4 spatial frequencies (3.0, 6.0, 12.0, and 18.0 cycles per degree [cpd]). Measurements of BSCVA under photopic (324 lux), scotopic (21 lux), and glare conditions were also performed using a Lighthouse Distance Visual Acuity Test, second edition (Lighthouse Low Vision Products). This uses a chart modified from the Early Treatment of Diabetic Retinopathy Study with Sloan letters. Glare conditions were simulated by a handheld brightness acuity test meter (Mentor) at its maximum setting. Visual acuity is reported as Snellen equivalents. The term best spectacle-corrected visual performance will be used to refer to these measurements. Topical proparacaine hydrochloride 0.5% (Ophthetic威), ofloxacin 0.3% (Ocuflox威), and diclofenac sodium 0.1% (Voltaren威) were administered immediately prior to the procedure. Appropriate alignment was achieved by marking the horizontal axis (0 to 180 degrees) with a surgical marking pen (Devon Industries), using a horizontal slit beam at the slitlamp with patients
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
VISUAL PERFORMANCE AFTER PARK
spatial frequencies (3.0, 6.0, 12.0, and 18.0 cpd). Contrast sensitivity measurements were made with the best spectacle correction in place. Each patient completed a survey rating adverse symptoms such as glare, halos, visual blurring, and color distortion. Each individual symptom was scored on a scale of 0 (none) to 5 (severe). This survey was conducted during the preoperative examination and at the 6, 12, and 24 month postoperative visits. Additionally, corneal clarity was assessed by slitlamp biomicroscopy and recorded on a scale of 0 to 5 by a single observer (E.E.M.). The SPSS Graduate Pack for Windows V6.1.3 (SPSS Inc.) was used to determine statistical significance. The analysis of predictors of poor visual performance was performed using a regression model. Differences were considered to be statistically significant when the P value was ⬍.05.
seated in the upright position. Patients were taken to the laser room and placed beneath the laser in the supine position. A wire speculum was placed in the eye being treated. The corneal epithelium was manually removed with a blunt PRK spatula. All patients had PARK at Stanford University Medical Center using the 193 nm Summit Apex Plus excimer laser delivered through a poly(methyl methacrylate) erodible mask system. Standard laser parameters were used in all cases. These included a repetition rate of 10 Hz, radiant exposure at the corneal plane of 180 mJ/cm2, and a 6.5 ⫻ 5.0 mm elliptical ablation zone. The ablation was delivered through an in-the-rail erodible mask provided by Summit Technology. Attempted correction was based on the manifest refraction. Postoperatively, a bandage contact lens was placed on the eye until the cornea had completely re-epithelialized. Postoperative eye medications included topical Ocuflox 4 times a day for 4 days or until the epithelium was healed and topical fluorometholone 0.1% (FML威) 4 times a day for 1 month and then 2 times a day for 1 month. Eyes were prospectively assessed at 1, 3, 6, 9, 12, 18, and 24 months following the procedure. All postoperative visits included a measurement of BSCVA under 3 standard conditions: scotopic (approximately 21 lux), photopic (approximately 324 lux), and glare. Contrast sensitivity measurements were performed 6, 12, and 24 months after the PARK procedure. Contrast sensitivity was evaluated using the CSV1000E (Vector Vision). The CSV-1000E uses a chart with vertical sine wave grating patterns. Each pattern represents an increase in contrast sensitivity determined by the manufacturer. The contrast sensitivity is recorded in log units. There are 9 contrast levels for each of the 4
Results The mean patient age was 41.4 years ⫾ 8.54 (SD) (range 26 to 58 years). Forty-nine eyes (52.7%) were in women and 44 (47.3%), in men. Prior to the 2 year follow-up, 24 eyes (25.8%) required retreatment. Seven eyes had LASIK enhancements and 17, PRK/PARK retreatment. These eyes were evaluated independently from the 2 year follow-up results. At the 2 year followup, 59 of the remaining eyes (71.1%) were available for analysis. Refractive results are shown in Table 1 and UCVA, in Table 2. Refractive and UCVA results for the subset of nonretreated eyes available at 1, 3, 6, 12, 18, and 24 months are shown in Table 3. All other visual acuity results are reported as best spectacle-corrected visual performance unless otherwise specified.
Table 1. Mean (⫾SD) sphere, cylinder, and SE at each of the designated intervals. Months Postop Preop
1
3
6
9
12
18
Sphere (D)
⫺3.88 ⫾ 1.83
⫺0.20 ⫾ 0.69
⫺0.33 ⫾ 0.75
⫺0.23 ⫾ 0.88
⫺0.22 ⫾ 0.88
⫺0.26 ⫾ 0.82
⫺0.16 ⫾ 0.92
0.05 ⫾ 0.86
Cylinder (D)
⫺2.19 ⫾ 0.90
⫺0.37 ⫾ 0.42
⫺0.41 ⫾ 0.45
⫺0.61 ⫾ 0.48
⫺0.64 ⫾ 0.52
⫺0.76 ⫾ 0.52
⫺0.75 ⫾ 0.52
⫺0.89 ⫾ 0.58
SE (D)
⫺4.98 ⫾ 1.80
⫺0.38 ⫾ 0.67
⫺0.53 ⫾ 0.68
⫺0.75 ⫾ 0.85
⫺0.54 ⫾ 0.81
⫺0.65 ⫾ 0.73
⫺0.53 ⫾ 0.80
⫺0.39 ⫾ 0.72
93
92
88
85
66
72
52
59
Number of eyes
24
SD ⫽ standard deviation; SE ⫽ spherical equivalent
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
1465
VISUAL PERFORMANCE AFTER PARK
Table 2. Uncorrected visual acuity results (number of eyes [%]) at each postoperative interval. Months Postoperative UCVA
1
3
6
9
12
18
24
ⱕ20/20
74 (80)
51 (58)
48 (56)
35 (53)
36 (50)
28 (54)
34 (58)
ⱕ20/40
87 (94)
78 (89)
70 (82)
57 (86)
67 (93)
48 (92)
56 (95)
Table 3. Refraction and UCVA results in subset of eyes not treated and available at specified months. Months Postop Preop
1
3
6
9
18
24
Sphere (D)
⫺3.78 ⫾ 1.90
⫺10.00 ⫾ 0.62
⫺0.14 ⫾ 0.64
⫺0.10 ⫾ 0.61
⫺0.71 ⫾ 0.77
⫺0.07 ⫾ 0.88
0.01 ⫾ 0.86
Cylinder (D)
⫺2.26 ⫾ 0.90
⫺0.33 ⫾ 0.43
⫺0.43 ⫾ 0.49
⫺0.58 ⫾ 0.47
⫺0.69 ⫾ 0.49
⫺0.78 ⫾ 0.51
⫺0.81 ⫾ 0.51
SE (D)
⫺4.94 ⫾ 1.91
⫺0.27 ⫾ 0.56
⫺0.36 ⫾ 0.48
⫺0.46 ⫾ 0.56
⫺0.52 ⫾ 0.65
⫺0.46 ⫾ 0.77
⫺0.39 ⫾ 0.74
ⱕ20/20
30 (60)
32 (64)
32 (64)
35 (70)
31 (62)
30 (60)
30 (60)
ⱕ20/40
48 (96)
44 (88)
45 (90)
46 (92)
50 (100)
47 (94)
47 (94)
Sphere, cylinder, and SE results are mean ⫾ SD; visual acuity results are number of eyes (%).
Contrast Sensitivity Function The spatial frequency and standard illuminance condition (scotopic and photopic) are specified for each figure. Tables 4 and 5 show the change in contrast sensitivity under both illuminance conditions and all measured spatial frequencies. Six months after the PARK treatment, the mean contrast sensitivity at all spatial frequencies under the photopic contrast measurements and at all low spatial frequencies (3.0 and 6.0 cpd) under the scotopic contrast measurements remained the same or decreased from the preoperative level. The decrease was statistically significant (P ⬍ .05) at the low spatial frequen-
cies under the photopic condition and at the lowest spatial frequency (3.0 cpd) under the scotopic condition. By 1 year, however, the mean contrast sensitivity at all spatial frequencies and all illuminance conditions had returned to the preoperative level (Tables 4 and 5). At all spatial frequencies and both illuminance conditions, the 12 and 24 month mean contrast sensitivities were higher than the 6 month mean contrast sensitivities. The mean contrast sensitivity at the 24 month visit was higher than the preoperative level under all conditions except the scotopic 3.0 cpd condition, which showed a statistically insignificant decrease in the mean contrast sensitivity of ⫺0.01 log units.
Table 4. Change in scotopic contrast sensitivity at each spatial
Table 5. Change in photopic contrast sensitivity at each spatial
frequency.
frequency.
Spatial Frequency (cpd)
24
Spatial Frequency (cpd)
6
12
6
12
24
3.0
⫺0.05 ⫾ 0.20
⫹0.02 ⫾ 0.26
⫺0.01 ⫾ 0.22
3.0
⫺0.07 ⫾ 0.19
⫹0.02 ⫾ 0.26
⫹0.01 ⫾ 0.21
6.0
0.00 ⫾ 0.31
⫹0.04 ⫾ 0.25
⫹0.08 ⫾ 0.23
6.0
⫺0.14 ⫾ 0.35
⫹0.04 ⫾ 0.25
⫹0.06 ⫾ 0.41
12.0
⫹0.18 ⫾ 0.65
⫹0.16 ⫾ 0.50
⫹0.21 ⫾ 0.55
18.0
⫹0.03 ⫾ 0.40
⫹0.20 ⫾ 0.40
⫹0.23 ⫾ 0.43
Months Postoperative
(P ⫽ .03)
(P ⬍ .01)
Change in contrast sensitivity (log units) is calculated from the preoperative contrast sensitivity measurement.
1466
Months Postoperative
(P ⬍ .01) 12.0
⫺0.04 ⫾ 0.51
⫹0.16 ⫾ 0.50
⫹0.09 ⫾ 0.44
18.0
0.00 ⫾ 0.40
⫹0.20 ⫾ 0.40
⫹0.13 ⫾ 0.35
Change in contrast sensitivity (log units) is calculated from the preoperative contrast sensivity measurement.
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
VISUAL PERFORMANCE AFTER PARK
Figure 1. (Haw) Mean change in BSCVA under scotopic conditions. Error bars represent 95% confidence interval.
Figure 2. (Haw) Mean change in BSCVA under photopic condi-
Figure 3. (Haw) Mean change in BSCVA under glare conditions. Error bars represent 95% confidence interval.
Figure 4. (Haw) Loss of BSCVA under scotopic conditions at
tions. Error bars represent 95% confidence interval.
24 months.
Figure 5. (Haw) Loss of BSCVA under photopic conditions at
Figure 6. (Haw) Loss of BSCVA under glare conditions at
24 months.
24 months.
Best Spectacle-Corrected Visual Performance Figures 1 to 3 show the mean change (lines reported in Snellen equivalents) in the best spectaclecorrected visual performance throughout the postoperative period under scotopic, photopic, and glare conditions. There was a mean decrease of ⫺0.56 ⫾ 1.19 lines 1 month after PARK under the scotopic condition (Figure 1). The scotopic best spectacle-corrected visual performance increased at every measured postoperative period. By 6 months, the best corrected visual performance reached the preoperative level. Figure 2 demonstrates a similar recovery in the photopic best corrected visual performance. At 1 month, there was a mean decrease of 0.60 ⫾ 1.10 lines. This also improved at every measured postoperative interval. The
photopic best corrected visual performance returned to the preoperative level by 9 months. As in the scotopic best corrected visual performance, recovery was more rapid between 1 and 3 months. Figure 3 demonstrates the mean change in the glare best corrected visual performance. At 1 month, there was a mean decline of ⫺0.33 ⫾ 1.32 lines. The glare best corrected visual performance returned to the preoperative level between 3 and 6 months. As in the scotopic and photopic best corrected visual performance conditions, there was an improvement at every measured postoperative visit. At the last follow-up, 0 eyes lost ⬎2 lines of best corrected visual performance under photopic or glare conditions, while 1 eye (1.7%) lost ⬎2 lines under scotopic conditions (Figures 4 to 6).
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
1467
VISUAL PERFORMANCE AFTER PARK
Visual Performance After Retreatment Twenty-four eyes (25.8%) had retreatment for residual refractive error. Insufficient results are available on the 7 eyes that had LASIK retreatment. Therefore, analysis is confined to the 17 eyes of 14 patients who had retreatment with PRK or PARK. The mean retreatment was 11.3 ⫾ 4.9 months (range 6 to 23 months) following the original PARK treatment. Eight eyes (47%) had PARK enhancement for residual compound myopic astigmatism; 9 eyes (53%) had PRK enhancement for residual spherical myopia. Prior to enhancement, the mean preoperative sphere was ⫺1.62 ⫾ 0.75 D (range ⫺0.75 to ⫺3.25 D), the mean preoperative astigmatism was ⫺0.93 ⫾ 0.64 D (range 0.00 to ⫺2.50 D), and the mean spherical equivalent was ⫺2.08 ⫾ 0.76 D (range ⫺1.13 to ⫺3.75 D). At the last follow-up (mean 12.9 months), the mean sphere was ⫺0.29 ⫾ 1.23 D, the mean astigmatism was ⫺0.88 ⫾ 0.40 D, and the mean spherical equivalent was – 0.74 ⫾ 1.27 D. The UCVA was 20/20 or better in 59% of eyes and 20/40 or better in 88%. There was no significant (P ⬎ .05) decrease in the mean contrast sensitivity from the pre-retreatment levels at all spatial frequencies and both illuminance conditions 6 months after retreatment. Similarly, no significant decrease was noted at the last follow-up. The mean contrast sensitivity under both illuminance conditions was not significantly lower than the mean preoperative levels. Four eyes (24%) experienced haze at the last postoperative visit following retreatment; 2 of them developed late-onset haze and regression. Among the 4 eyes, 2 lost 2 lines of best corrected visual performance under glare conditions, and 1 lost 2 lines of best corrected visual performance under both scotopic and photopic conditions. The worst BSCVA was 20/32 under any of these conditions. No eye lost more than
2 lines of best corrected visual acuity under scotopic, photopic, or glare conditions. Corneal Haze Corneal haze was graded on a scale of 0 to 5 at the slitlamp. The mean level of corneal haze increased progressively up to the 12 month follow-up. The corneal haze diminished at the 18 and 24 month follow-ups. The mean corneal haze score decreased 47% from its peak value of 0.32 ⫾ 0.80 at 12 months to 0.17 ⫾ 0.46 at 24 months. Eight eyes (13.5%) had trace to mild haze and no eye had moderate or severe haze at the 24 month follow-up. At the point of maximum corneal haze (12 months), 11 eyes (15.3%) demonstrated grade 1 to grade 3 haze. These eyes demonstrated a loss of best spectacle-corrected visual performance of ⫺0.64 ⫾ 1.50 lines under scotopic conditions, ⫺0.27 ⫾ 0.90 lines under photopic conditions, and ⫺0.09 ⫾ 0.94 lines under glare conditions. This compared adversely with the mean increase in best spectacle-corrected visual performance in eyes without haze in which the respective values were ⫹0.49 ⫾ 1.09 (P ⫽ .09), ⫹0.26 ⫾ 1.21 (P ⫽ .62), and ⫹0.82 ⫾ 1.16 (P ⫽ .24). At 24 months, the mean best corrected visual performances under scotopic and glare conditions remained lower in eyes with haze than in those without haze (Table 6). However, these values were not statistically significant. Eyes with haze had a tendency to demonstrate a more pronounced decrease in the contrast sensitivity function curve under lower spatial frequencies (3.0 and 6.0 cpd) than eyes with no corneal haze at the 6 month visit. However, this was only statistically significant for the 6.0 cpd data point under photopic conditions (P ⫽ .02). By 12 months, there was no statistically significant difference between eyes with or without haze in the contrast sensitivity function curve at all measured spatial
Table 6. Change in best corrected visual performance in eyes with and without haze under each condition. 6 Months Condition Scotopic Photopic Glare
12 Months
24 Months
ⴙ Haze
ⴚ Haze
ⴙ Haze
ⴚ Haze
ⴙ Haze
ⴚ Haze
⫺0.13 ⫾ 0.74
⫹0.10 ⫾ 1.04
⫺0.65 ⫾ 1.50
⫹0.49 ⫾ 1.09
⫹0.25 ⫾ 0.89
⫹0.64 ⫾ 0.94
0.00 ⫾ 1.00
⫺0.29 ⫾ 1.03
⫺0.27 ⫾ 0.90
⫺0.27 ⫾ 0.90
⫹0.50 ⫾ 0.76
⫹0.48 ⫾ 0.84
⫹0.19 ⫾ 0.98
⫹0.33 ⫾ 1.24
⫺0.09 ⫾ 0.94
⫹0.82 ⫾ 1.16
⫹0.25 ⫾ 0.85
⫹0.78 ⫾ 0.99
Mean change in best corrected visual performance recorded in change of lines (Snellen equivalents) from the preoperative value.
1468
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
VISUAL PERFORMANCE AFTER PARK
Figure 7. (Haw) Change in contrast sensitivity in eyes with and without haze under scotopic and photopic conditions.
frequencies and illuminance conditions. At the last follow-up at 24 months, there was a slight tendency toward a lower contrast sensitivity in eyes with haze than in eyes without haze at all but the highest spatial frequency (18.0 cpd) (Figure 7). At this postoperative interval, only the photopic 3.0 cpd (P ⬍ .01) data point demonstrated a statistically significant decrease between eyes with and without haze. Survey of Symptoms The increases in the mean glare, visual blurring, and color distortion scores were not statistically significant (P ⬎ .05) at the last postoperative visit. Despite an improvement in the halo score from months 18 to 24, there was a statistically significant increase in the halos from the preoperative level. The preoperative score of 0.64 ⫾ 0.98 increased to 1.09 ⫾ 1.20 at the 24 month follow up (P ⫽ .04). Predictors of Poor Visual Performance A regression model was used to evaluate potential predictors for the relative decline in the visual performance within the first 6 months, prior to the exclusion of retreatment eyes. A regression was also used to determine whether a change in the glare score, halo score, visual blurring score, and color distortion score predicted declines in the visual performance over the same interval. The effect of the level of spherical equivalent correction and the level of astigmatic correction on the contrast sensitivity function curve was specifically evaluated. The photopic contrast sensitivity function curve was not significantly affected (P ⬎ .05) by either preoperative variable. However, eyes with high preoperative spherical equivalents were predictive of an elevated scotopic con-
trast threshold at the extreme spatial frequencies (3.0 cpd [P ⫽ .02] and 18.0 cpd [P ⫽ .03]) 6 months after PARK. The best corrected visual performance under glare and photopic conditions at 1, 3, and 6 months was not affected by the level of preoperative spherical equivalent or the level of preoperative astigmatism (P ⬎ .05). However, poor best corrected visual performance under scotopic conditions at 1 month was predicted by the level of preoperative astigmatism (P ⬍ .05). This effect was only temporary and by postoperative month 3, there was no predictive effect of preoperative astigmatism (P ⬎ .05). The changes in the glare, halo, visual blurring, and color distortion scores did not predict changes in either the contrast sensitivity function curve under scotopic and photopic conditions or the best corrected visual performance under glare, scotopic, and photopic conditions (P ⬎ .05). Vector analysis performed at the 24 month follow-up demonstrated no predictive effect of surgically induced astigmatism, target induced astigmatism, mean angle of error, difference vector, or mean magnitude of error on the loss of BSCVA under glare, photopic, or scotopic conditions. A higher mean angle of error was weakly predictive of lower contrast sensitivity under scotopic illuminance at 12.0 cpd (P ⫽ .047). No other vector analysis variable was predictive of poor contrast sensitivity (P ⬎ .05).
Discussion There is substantial evidence that contrast sensitivity is adversely affected following refractive surgical procedures. Some studies demonstrate a measurable decrease in the contrast sensitivity that returns to normal 6 months to 1 year after the refractive surgical procedure.8,14,16 However, other studies suggest that visual function on psychophysical tests may still be reduced at or beyond 1 year.2–3,9,12 In our cohort, there was a tendency for a relative decline in the contrast sensitivity function curve in the early postoperative period under both scotopic and photopic conditions. This was most pronounced under photopic illuminance conditions, as well as at the low spatial frequencies at the 6 month visit. The recovery of mean contrast sensitivity function was most profound at the higher spatial frequencies and was demonstrated by a
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
1469
VISUAL PERFORMANCE AFTER PARK
tendency toward improvement in contrast sensitivity at the 6 month visit and by a statistically significant increase in contrast sensitivity at 12 month. By 12 months, the mean contrast sensitivity had returned to normal at all spatial frequencies and under both illuminance conditions. The contrast sensitivity remained stable at the 24 month visit. We noted a tendency toward decreased performance on the contrast sensitivity function curve under photopic conditions at the 6 month follow-up. Although the physiologic pupillary dilation that occurs during scotopic illumination may allow optical irregularities outside the photopic optical zone to affect the contrast perception, this may have been offset by the background photopic illuminance, which may inherently decrease the contrast of the black letters on the back-lighted contrast chart. However, the best spectaclecorrected visual performance under scotopic and photopic conditions was less sensitive to this difference. Figures 1 and 2 mirrored each other during the postoperative time course. The general improvement in contrast from preoperatively may not be due to a “real” increase in contrast sensitivity following PARK but rather a “learning” effect. This phenomenon is well documented in patients taking difficult examinations such as automated visual field tests in which an artifactual deficit may improve on a repeat examination. This may also occur following the measurement of contrast sensitivity. Patients may “learn” to improve their performance on subsequent tests by repeat examinations. Since the patients had not had previous trials at measuring the contrast sensitivity, the preoperative level may be expected to be the lowest. Nevertheless, it is difficult to determine to what degree (if any) this phenomenon accounted for our results. Certainly, age-matched control groups that had not had PARK but had had repeat contrast sensitivity testing would have been useful in determining the degree to which this effect accounted for our observed results. In addition, we introduced a selection bias by independently analyzing a significant number of eyes that were retreated (26%). We do not know whether this selection bias may have contributed to the observed trends in visual function since the retreatment eyes did not experience a statistically significant decrease in their best spectacle-corrected contrast sensitivity function. 1470
The best corrected visual performance under scotopic, photopic, and glare conditions significantly decreased 1 month after PARK. However, there was a steady and progressive improvement in the best corrected visual performance under all 3 conditions at every measurable postoperative visit from 1 to 24 months. The recovery was most rapid in the early postoperative period between 1 and 3 months. Glare recovery was more rapid, with full recovery occurring at approximately 3 months rather than 6 to 9 months with the photopic and scotopic best corrected visual performances. The mean improvement in scotopic, photopic, and glare best corrected visual performances remained stable through the 24 month follow-up. Our regression analysis suggests that the level of preoperative spherical equivalent and the level of preoperative astigmatism were not predictive of a poor outcome on the photopic contrast sensitivity function curve 6 months after PARK. Although the level of preoperative astigmatism did not adversely affect the scotopic contrast sensitivity function curve, it was affected by the degree of preoperative spherical equivalent at the extreme spatial frequencies (3.0 and 18.0 cpd) during the same interval. This is not entirely surprising as a higher preoperative spherical equivalent corresponds to a larger ablation depth and a steeper transition between the unablated and ablated zone (6.5 ⫻ 5.0 mm ablation zone in our trial). This steeper transition would be expected to play a larger role under scotopic (physiologic pupillary dilation) conditions. It has been suggested that the Vector Vision CSV-1000 system used to measure contrast sensitivity in our study may unmask aberrations from the transition zone of ablated and unablated corneas in PRK.15 Similarly, the attempted level of astigmatic correction was predictive of an early poor outcome in the scotopic best corrected visual performance at 1 month. However, this had resolved by the 3 month postoperative visit. A more complicated astigmatic ablation may result in larger contrast sensitivity declines than a spherical ablation. However, this may be offset by the use of erodible mask technology, which results in a smoother ablation profile that would be expected to diminish the effect of optical aberrations resulting from an irregular ablation. In a study of 30 eyes that had RK and 30 eyes that had PRK, contrast sensitivity and glare disability testing
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
VISUAL PERFORMANCE AFTER PARK
did not accurately reflect patients’ subjective assessment of their visual performance in daily life.14 Our study also did not find the degree of subjective symptoms of glare, halo, blurring, or color distortion to accurately predict the declines in contrast sensitivity function or best corrected visual performance under scotopic, photopic, or glare conditions. Verdon and coauthors3 reported 18 patients who had PRK for –5.08 ⫾ 1.63 D of myopia. The authors concluded that there was no notable relationship between corneal haze and any visual performance measure. However, they cautioned that this might have resulted from the coarse grading scale used to assess haze, a restricted range of haze values, and a homogeneous and small sample size. In our study, eyes with corneal haze demonstrated a trend toward decreased contrast sensitivity at the lower spatial frequencies under both scotopic and photopic illuminance conditions at the earliest (6 month) measurement. Despite the tendency for improvement in the mean contrast sensitivity of treated eyes from the 6 month interval, eyes with persistent haze at the 24 month interval retained a lower mean contrast sensitivity at the lower (3.0 to 12.0 cpd) spatial frequencies. These eyes also demonstrated a lower best corrected visual performance than eyes without haze under scotopic and glare conditions by the 24 month followup. However, the photopic best corrected visual performance had returned to normal by this time. Retreatment with PRK or PARK did not adversely affect contrast sensitivity under scotopic or photopic illuminance conditions at any time during the postretreatment period. When compared with the mean preoperative contrast sensitivity prior to the original PARK procedure, there was also no statistically significant decrease in the contrast sensitivity under both illuminance conditions. However, best spectacle-corrected visual performance was adversely effected by late-onset haze in 2 retreated eyes. Our results appear to suggest a measurable decline in many of our psychophysical measurements with early, rapid improvements and return to the preoperative state by 6 months to 1 year following the treatment of compound myopic astigmatism with PARK. Possible predictors of early poor visual performance under scotopic conditions include attempted level of correction for spherical equivalent and astigmatism. Corneal haze may also adversely affect the best spectacle-corrected visual
performance under scotopic and glare conditions. We found a poor correlation between the patient’s subjective grading of symptoms with the patient’s objective performance on our psychophysical measurements. The psychophysical assessments tended to remain stable up to 24 postoperative months.
References 1. Ambrosio G, Cennamo G, De Marco R, et al. Visual function before and after phtorefractive keratectomy for myopia. J Refract Corneal Surg 1994; 10:129 –136 2. Butuner Z, Elliott DB, Gimbel HV, Slimmon S. Visual function one year after excimer laser photorefractive keratectomy. Refract Corneal Surg 1994; 10:625– 630 3. Verdon W, Bullimore M, Maloney R. Visual performance after photorefractive keratectomy, a prospective study. Arch Ophthalmol 1996; 114:1465–1472 4. Dutt S, Steinert RF, Raizman MB, Puliafito CA. Oneyear results of excimer laser photorefractive keratectomy for low to moderate myopia. Arch Ophthalmol 1994; 112:1427–1436 5. Lohmann CP, Gartry DS, Muir MK, et al. Corneal haze after excimer laser refractive surgery: objective measurements and functional implications. Eur J Ophthalmol 1991; 1:173–180 6. O’Brart DPS, Lohmann CP, Fitzke FW, et al. Disturbances in night vision after excimer laser photorefractive keratectomy. Eye 1994; 8:46 –51 7. O’Brart DPS, Lohmann CP, Fitzke FW, et al. Night vision after excimer laser photorefractive keratectomy: haze and halos. Eur J Ophthalmol 1994; 4:43–51 8. Hodkin MJ, Lemos MM, McDonald MB, et al. Near vision contrast sensitivity after photorefractive keratectomy. J Cataract Refract Surg 1997; 23:192–195 9. Niesen U, Businger U, Hartmann P, et al. Glare sensitivity and visual acuity after excimer laser photorefractive keratectomy for myopia. Br J Ophthalmol 1997; 81: 136 –140 10. Corbett MC, Prydal JI, Verma S, et al. An in vivo investigation of the structures responsible for corneal haze after photorefractive keratectomy and their effect on visual function. Ophthalmology 1996; 103:1366 –1380 11. Deitz MR, Piebenga LW, Matta CS, et al. Ablation zone centration after photorefractive keratectomy and its effect on visual outcome. J Cataract Refract Surg 1996; 22: 696 –701 12. Niesen UM, Businger U, Schipper I. Disability glare after excimer laser photorefractive keratectomy for myopia. J Refract Surg 1996; 12:S267–S268 13. Ficker LA, Bates AK, Steele AD McG, et al. Excimer laser photorefractive keratectomy for myopia: 12 month follow-up. Eye 1993; 7:617– 624 14. Ghaith AA, Daniel J, Stulting RD, et al. Contrast sensi-
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000
1471
VISUAL PERFORMANCE AFTER PARK
tivity and glare disability after radial keratotomy and photorefractive keratectomy. Arch Ophthalmol 1998; 116: 12–18 15. Wachler BSB, Durrie DS, Assil KK, Krueger RR. Role of clearance and treatment zones in contrast sensitivity: significance in refractive surgery. J Cataract Refract Surg 1999; 25:16 –23 16. Wang Z, Chen J, Yang B. Comparison of laser in situ keratomileusis and photorefractive keratectomy to cor-
1472
rect myopia from ⫺1.25 to ⫺6.00 diopters. J Refract Surg 1997; 13:528 –534 17. Carones F, Brancato R, Morico A, et al. Compound myopic astigmatism correction using a mask in-the-rail excimer laser delivery system. Preliminary results. Eur J Ophthalmol 1996; 6:221–233 18. Maloney RK, Friedman M, Harmon T, et al. A prototype erodible mask delivery system for the excimer laser. Ophthalmology 1993; 100:542–549
J CATARACT REFRACT SURG—VOL 26, OCTOBER 2000