Excimer Laser Photorefractive Keratectomy l8--month Follow,up ...
David S. Gartry, MD, FRCS, FCOphth, Malcolm G. Kerr Muir, MD, FRCS, FCOphth, John Marshall, PhD Purpose: This study, comprising 120 patients (follow-up, 12 to 22 months; mean, 18 months), was designed to evaluate the efficacy and safety of excimer laser photorefractive keratectomy. Methods: Ablation zone diameter (4 mm), ablation rate (0.22 ~m/pulse), fluence (180 mJ/cm2), and frequency (10Hz) were kept constant, and all patients used a standard topical corticosteroid regimen. Results: Overcorrection was followed by regression to a stable refraction by 4 months. At 1 year, 95% and 70%, respectively, of patients undergoing -2.00 diopters (D) and -3.00 D corrections and 40% and 20% of those undergoing -6.00 D and -7.00 D corrections were within ±1.00 D of intended refraction. Anterior stromal "haze," maximal at 6 months and diminishing thereafter, was detected in 110 patients (92%). However, best-corrected Snellen visual acuity was reduced in only 22 (18%). Good correlation existed between haze and regression (r = 0.68). Night "halos," due to the relatively small 4-mm diameter ablation zone, were reported by 94 patients (78%) in the early postoperative period, and, at 1 year, 12 (10%) declined treatment of the other eye because of persistence of this problem. Conclusion: Marked individual variation was found after photorefractive keratectomy. However, in low myopia, predictability is "fair," and the procedure is safe. The authors conclude that excimer laser photorefractive keratectomy holds considerable promise for refractive surgery in the future. Ophthalmology 1992;99: 1209-1219
The potential of the argon fluoride excimer laser in ophthalmic surgery is based on two characteristic features. First, operating at a wavelength of 193 nm, corneal tissue
Originally received: January 22, 1992. Revision accepted: March 16, 1992. From the Department of Ophthalmology, St Thomas' Hospital, London. Presented at the American Academy of Ophthalmology Annual Meeting, Anaheim, October 1991. Supported by the Iris Fund for Prevention of Blindness, London, England, including a research fellowship sponsored by the Iris Fund for Prevention of Blindness awarded to Dr. Gartry. Professor Marshall is a consultant for Summit Technology. Drs. Gartry and Kerr Muir have no proprietary interest in the subject matter of this article. Reprint requests to David Gartry, MD, FRCS, FCOphth, Moorfields Eye Hospital, City Road, London EC I V 2PD, England.
can be removed with a high degree of precision with virtually no damage to adjacent unexposed areas. 1-3 Second, the beam has a relatively large diameter (5 to 6 mm) and can be configured in any cross-sectional shape. When this beam passes through a computer-controlled expanding iris diaphragm, the corneal surface can be directly flattened to effect a reduction in myopia. This no-touch procedure has been termed photorefractive keratectomy (PRK).4 Computer-generated algorithms relating the diameter and depth of ablation needed for a given refractive change were devised based on the average amount of tissue removed per pulse. 5 Using these algorithms, a -7.00-diopter (D) correction can be achieved with removal of only 44 ~m of tissue (less than 10% of the axial corneal thickness) when a 4-mm diameter beam is used. The acute change and the healing response were studied in animals using experimental systems,6-1 3 and several ophthalmic excimer lasers are currently undergoing clinical evaluation. These
1209
Ophthalmology
Volume 99, Number 8, August 1992 Table 1. Patient Selection Criteria
systems were first used to treat superficial corneal pathology and the high degree of precision of tissue removal was confirmed. 14 More recently, treatment of refractive error, principally myopia, has been addressed. 15 - 23 This report presents the results of the first sighted eye cohort to undergo PRK in the United Kingdom. The study, which began in November 1989 at St. Thomas' Hospital, London, comprised 120 eyes (120 patients) with followup from 12 to 22 months (mean, 18 months).
Astigmatism Visual acuity
Materials and Methods
Age Contact lens history
Parameter Refraction
Laser A Summit Technology UV200 Excimed laser (Waltham, MA) with an emission wavelength of 193 nm, a fixed pulse repetition rate of 10Hz, and a radiant exposure of 180 mJ/cm 2 at the cornea was used. Early laboratory studies indicated that at this energy level approximately 0.22 !lm of corneal tissue was removed per pulse. To facilitate comparison among patients, this standard value was programmed in all procedures (although lasers manufactured since the beginning of the current study [November 1989] have incorporated a fixed value of 0.25 !lm/pulse in the algorithm). In addition, a standard ablation zone diameter of 4 mm (the maximum available at the beginning of the study) was selected in each case. There was no effluent removal system and no eye stabilizing system. The patient was required to fixate a target light within the laser aperture for the duration of the procedure (14 seconds on average). Human Studies After 2 years of experience in the treatment of corneal pathology, 14 Ethical Committee approval was given for a two-phase study; first, in blind eyes to allow assessment of the safety of the procedure and consolidate the surgical technique; and second, given satisfactory results, in sighted eyes to determine predictability and stability of the induced refractive change. The Blind Eye Study and preliminary sighted eye results have been reported in detail elsewhere. 15 Patient Selection Criteria Selection criteria are summarized in Table 1. Prospective patients were sent background information, in which the research nature of the project was emphasized. After questionnaire and interview stages, suitable patients underwent a detailed preoperative examination (Table 2). Study Design Depending on preoperative refraction, patients were allocated to I of 6 treatment groups (-2.00 D through to -7.00 D in whole diopter steps) with 20 patients per group. Within a group, all patients had an identical PRK treat-
1210
Ophthalmic history Occupation
Motivation
Informed consent
Criterion Patients with myopia (all degrees) with stable refraction; anisometropic patients included Less than 1.00 diopter correction Best-corrected visual acuity 20/30 or greater in both eyes Lower limit 24 years Contact lens wearers preferred due to likelihood of induced anisometropia; mainly soft lens wearers; patients with evidence of fluctuating refraction due to corneal distortion were excluded Patients with a history of ocular disease were excluded Individuals with exacting vocational visual requirements were excluded (e.g., pilots, professional drivers) Prospective patients were asked to state their reasons for wishing to be included in the trial; only those with realistic expectations were included Only those patients demonstrating a full understanding of the research nature of the trial were included; a detailed informed consent was signed prior to the surgery
ment (i.e., the same diopter correction was attempted). This allowed assessment of healing of identical ablation depths within each group. The range of preoperative myopia was -1.50 D to -17.50 D (spherical equivalent). Because the maximum treatment programmed was -7.00 D, and approximately one fifth of the cohort had myopia in excess of this, only partial correction was attempted in these patients. Anisometropic patients also were included and therefore emmetropia was not the aim in every case. To evaluate individual variation, all variables within our control were kept constant among patients: ablation rate (0.22 !lm/pulse), diameter of ablation zone (4 mm), surgical technique and postoperative treatment regimen (high-dose topical corticosteroids for 3 months; see below). Postoperative Management and Monitoring Immediately after the procedure was completed, slit-lamp examination was performed and the cornea was photographed (Fig 1). Oral analgesics (dextropropoxyphene/paracetamol [Distalgesic] X 2) were given immediately, and 4 tablets were supplied for home use. Topical cyclopentolate 1%,
Gartry et al . Excimer Laser Photorefractive Keratectomy Table 2. Schedule of Investigations and Follow-up Preoperative 1/7 5/7 2/52 4/52 6/52 8/52 10/52 3/12 4/12 6/12 9/12 1yr 18/12 Investigation Refraction Retinoscopy Autorefraction Keratometry Photography Halo measurement Endothelial specular microscopy Slit-lamp lOP Contrast sensitivity Videoimaging lOP
=
+ + + + +
+.... + + + +
+ +
+ +
+ +
+ +
+ +
+ +
+ + +
+ + +
+ + +
+ + +
+ + +
+ + +
+
+
+ +
+
+
+
+ +
+ +
+ + + + + + + + + + +
+ +
+ +
+ +
+ +
+ +
+ + +
+ + +
+ + +
+ + +
+ + +
+
+
+
+ + +
+
+ +
+
+ +
intraocular pressure.
phenylephrine 10%, homatropine 2%, and betamethasone with neomycin ointment were instilled, and the eyelids were taped and padded for 16 hours. The accepted role of topical corticosteroids in treatment of corneal disease, their capacity to modulate experimental wound healing (mainly linear incisions in rabbits), and limited experimental evidence that their use also might minimize corneal scarring after PRK9 resulted in Ethical Committee constraints such that a fixed regimen of topical dexamethasone 0.1 % 5 times per day for 2 months, tapering over the third month, was used in all patients (irrespective of refractive status or corneal clarity). This standardized regimen also ensured that comparisons could be made
Figure 1. Corneal appearance immediately after a -6.00 0 PRK procedure. The concentric ring pattern provides evidence of steady fixation and also can be used to assess the centration of the ablated zone in relation to the optical axis. The irregular light reflex beyond the ablation zone is the edge of the area of epithelial debridement.
among patients. (As a result of the current study, however, permission was granted for a randomized, double-masked trial with and without topical corticosteroids. The results of this trial will be reported soon). Topical chloramphenicol 0.5%, 4 times per day, also was used for the first 2 weeks. The follow-up intervals and assessment procedures at each stage are summarized in Table 2. During subjective refraction, reliance was placed on fogging techniques, in particular the + 1.00 D blur test, to ensure that an accurate end point was reached consistently. This test assumes that when a further, unnecessary + 1.00 D is added, the patient will be "fogged" to approximately 20/60. If, with this extra + 1.00 D, Snellen acuity is significantly better than 20/ 60, then too much minus has been prescribed and the spherical component can be revised accordingly. It was not possible to use cycloplegic refraction because of unwanted aberration effects from the paracentral cornea. Corneal clarity was subjectively graded on a scale from 0 to 4, as described in detail previously. IS Night halo effects, which were found to be due to the relatively small diameter of the ablation zone (4 mm), could be eliminated either by miosis or by interposing an appropriate negative lens. By using the internally illuminated standard Snellen visual acuity chart against a dark background, a subjective assessment of the magnitude of the halo effect could be made. After accurate refraction at each postoperative visit, patients were asked to estimate the width of the halo in comparison to the width of the Snellen chart. Increasing powers of negative lenses were then placed in the trial frame until the halo was eliminated. The minimum power of minus lens required to eliminate the halo was recorded. To determine its effect on the magnitude of the halo, an estimate of pupil size for each patient was made using a hand-held comparator (under the same clinical conditions used for measurement of the halo with the Snellen chart).
1211
Ophthalmology
Volume 99, Number 8, August 1992
Results Approximately 2 hours after the procedure, all patients reported pain r~n~ng from mild discomfort to moderately sev~re. The maJonty slept well, although a significant minonty woke on one or more occasions. Most used all 6 Di~talgesic tablets supplemented with proprietary analgesICS. The pain settled within 24 hours and all patients reported that the eye felt normal after 4 days. By the fifth day, re-epithelialization was complete. Unaided vision was reported to be considerably improved by day 4, and by week 1 accurate subjective refraction was possible. Best spectacle-corrected visual acuity returned to preoperative levels by week 2. On direct questioning, no patient reported diurnal variation. Postoperative Refraction A series of graphs relating mean change in refraction to follow-up time is shown in Figures 2A to F. All patients had. a significant decrease in myopia in the first postoperatIve week. A mean overcorrection was noted propor~ional to the magnitude of the attempted correction (e.g., In the -3.00 0 treatment group the mean change was 5.500, while in the -7.00 D group it was 10.25 D). This overcorrection reduced progressively between the second and twelfth postoperative weeks. Further regression occurred in some individuals in all groups but was more common in the -6.00 D and -7.00 0 groups (Figs 2E and F). In the -7.00 D group, the degree of immediate postoperative change also was related to the amount of preoperative myopia. In analyzing data, patients in this group were subdivided into those whose preoperative refractive error was less than -10.00 D (mean, -7.600; n 10) and those whose preoperative refractive error was greater than -10.00 D (mean, -12.94 D; n = 10) (Fig 3). The mean change at the first postoperative visit for the former group was 9.00 D (an overcorrection of 2.00 D), and for the latter group 12.25 D (an overcorrection of 5.25 D). The -2.00 D, -3.00 D, -4.000, and -5.00 D groups demonstrated minimal change in mean refraction after the third month (Figs 2A to D). In contrast, in the -7.00 D group, a more marked regression toward the preoperative level occurred throughout the first 6 months after which mean refraction stabilized (Fig 2F). Although the mean change in refraction achieved at 3 months in all groups was close to that intended, there was considerable individual variation (Fig 4). The refraction results at 18 months are summarized in Table 3 and the unaided visual acuity, in those patients in whom a full correction was attempted, are summarized in Table 4. There was a ~igh direct correlation between keratometry and refractIon. No statistically significant change in astigmatism was fo~nd when compared with the preoperative amount (twot~Iled Student's t test, P> 0.05). Retinoscopy was relatIvely easy to perform postoperatively, and a well-defined central zone of altered reflex was clearly visible. The central "with" movement was surrounded by an annulus of "against" movement. Retinoscopy was most useful in the
1212
Table 3. Accuracy of Photorefractive Keratectomy (Follow-up 1 Year) Patients within Stated Tolerances
Treatment Group (D)
±O.50D
±1.00D
Patients Achieving >1/2 Attempted Correction (%)
-2.00 -3.00 -4.00 -5.00 -6.00 -7.00
75% 45% 20% 30% 25% 5%
95% 70% 40% 50% 40% 20%
95% 90% 65% 80% 60% 70%
early postoperative period to confirm the amount of overcorrection present, and to minimize the tendency to accommodate before subjective refraction. Visual Acuity and Anterior Stromal Haze Correlations At 1 year, 98 patients (82%) had best spectacle-corrected visual acuity equal to or better than preoperative visual acuity. Eighteen patients (15%) lost between liz and 1 line of Snellen acuity, while 4 patients (3%) lost 2 lines (Table 5). These 4 patients had undergone either -6.00 0 or -7.00 ~ procedures (ablation depths of 39 and 44 JLm, ~esp~ctIvely). As expected, the irregularities of healing epIthelIum caused a slight reduction in clarity in the first postoperative week. After this, the cornea was clear until the 3- to 6-week stage, when a faint, diffuse subepithelial stromal haze appeared that increased to a maximum by 6 months. This change in corneal clarity (haze) was detected in 110 patients (92%) by the third month. The magnitude and time course of this haze, as determined by our subjective grading method, are shown in Figure 5. There was only limited statistical correlation of haze with age (r = -0.32) and poor correlation with ablation depth (r = 0.22) despite the clinical impression that older patients demonstrated less haze, and higher order corrections (deeper ablations) resulted in more marked haze. There was good correlation, however, between haze and the regression of refraction (r values for the-2.00 D to -7.00 D groups = 0.49, 0.75, 0.83, 0.36, 0.86, and 0.76, respectively; mean correlation, 0.68). In addition in those patients in whom correction should have prod~ced emm.etropi~ (n = 66), unaided vision was highly correlated WIth reSIdual ametropia at 1 year (r = 0.94). Side Effects ~able 5 identifies complications and their percentage inCIdence. The one case of significant decentration of the ablation zone occurred in the third patient to be treated. Fixation was maintained, but viewing angle altered when the patient moved steadily away from the surgeon. This could have been avoided by interrupting the treatment
Gartry et al . Excimer Laser Photorefractive Keratectomy Change in Refraction following -2.00D PRK (Ablation depth
Change in Refraction following -3.000 PRK
= 17um) 14
8 mean preoperative refraction
Ci
r::
0
-2.10D
6
(n
= 20)
«J
.= Q)
CI
2
r::
o
10
-
mean preoperative refraction
.........
--~
n = 20
t1
.&:.
. ...
0 0
8
Q)
a:
(J
A
-
-3.86D
«J
4
«J
12
r::
o
0
.........Q) a:
e
(Ablation depth = 23 urn) ~----_.-----,----_r----_r----_r----_..
3
6
.=
r r
1 9
12
f
15
6
Q)
4
CI
r::
.«J
.&:.
2
(J
o
18
B
Months
3
6
9
12
15
18
Months Change in Refraction following -5.000 PRK
Change in Refraction following -4.000 PRK (Ablation depth
o
= 28um)
(Ablation depth = 33um)
14
12 mean preoperative refra c tion
Cl
r:: 0 .;:;
12
mean preoperative refraction -4.00D
10
(n = 20)
-a:
8
r:: Q)
CI
r::
.=
6
Q)
4
(J
.&:.
2
(J
0 0
6
9
12
15
0
D
r-r----,-----,----_.-----,----_r----_r~
-7 .00D
10
(n= 20)
.&:.
12
15
18
mean preoperative refracti on Cl
r::
.2
0«J
1\
8
-
6
------------------------
-a:....
12
-10 .26D
(n = 20 )
10 8
Q)
.=
6
CI
4
Q)
4
r::
r::
(J
9
(Ablation depth = 44um)
14
CI «J
6
0
Change in refraction following -7.000 PRK
c: Q)
2
Change in Refraction following -6.000 PRK
«J ....
Gi a:
---------------r------
f--4
Months
r::
.g
6
Months
mean preoperative refraction Cl
8
0
18
(Ablation depth = 39um) 12
= 20)
r::
«J
C
(n
CI
«J
.&:.
0
.t: Q) a:
Q)
-7 .13D
10
r::
0«J
0
«J ....
Cl
«J
.&:.
2
(J
0
E
0
3
6
9
Months
12
15
18
2
0
F
0
3
6
9
12
15
18
Months
Figure 2A to F. Graphs relating mean change in refraction to time. Initial overcorrection in all cases is followed by regression to a plateau at between 3 and 6 months. The error bars are ±1 standard deviation. (Minimum follow-up for all patients is 12 months; maximum, 22 months; mean, 18 months).
1213
Ophthalmology
Volume 99, Number 8, August 1992
until the patient was repositioned. This patient experienced a greater amount of halo effect (vide infra) below lights, which gave rise to difficulty driving at night. Fourteen patients (12%) had a measurable rise in intraocular pressure (lOP) detected at the 4- to 6-week stage (topical corticosteroid responders). Of these, 3 reported transient loss of vision on rising rapidly from a chair and were found to have lOP in excess of 45 mmHg. Topical corticosteroids were discontinued, and lOP returned to normal within 2 weeks. Those with lOP in excess of 30 mmHg were treated with topical timolol maleate 0.25% twice a day for 1 month.
Individual Variation following PRK ( n = 120 at one year foll o w-up)
e z
o i= u W
~
o U o >
W
Desired
7
,........ ........ ..-- Correction 0 ....... / /
6 o
8
5
e..,.................
4
...... o
;"//
0...... . . . .
/"'"0
~
....... - B
Mean
0
g
iii
-4
-5
...--
Correction
Achieved
3
w
J:
u
Halo Effect (Low Illumination Positive Spherical Aberration) On direct questioning, 94 patients (78%) reported halos around sources oflight at night in the early postoperative period. The magnitude of this halo effect was highly directly correlated with pupil size (r = 0.79) and the amount of the induced refractive change (r = 0 .71) (Figs 6A and B). Those patients with greater overcorrection in the early postoperative phase experienced the most marked halo effect. With regression in refraction in the first 3 months subjective appraisal of the halo effect and a reduction in the power of the negative lens required to eliminate it indicated that the magnitude of the halo became less with time. However, at 1 year, 12 patients (10%) found that the effect persisted and was severe enough to interfere with night driving. They therefore declined to have the other eye treated. The effect was found to be unrelated to the lllagnitude and time course of the anterior stromal haze.
Change in Refraction following -7.00D PRK 14
0
'"
C>
0
4
2
C> Gl
Gl
Z
60 GlQ.'I;j
David S. Gartry, MD, FRCS, FCOphth, Malcolm G. Kerr Muir, MD, FRCS, FCOphth, John Marshall, PhD Purpose: This study, comprising 120 patients (follow-up, 12 to 22 months; mean, 18 months), was designed to evaluate the efficacy and safety of excimer laser photorefractive keratectomy. Methods: Ablation zone diameter (4 mm), ablation rate (0.22 ~m/pulse), fluence (180 mJ/cm2), and frequency (10Hz) were kept constant, and all patients used a standard topical corticosteroid regimen. Results: Overcorrection was followed by regression to a stable refraction by 4 months. At 1 year, 95% and 70%, respectively, of patients undergoing -2.00 diopters (D) and -3.00 D corrections and 40% and 20% of those undergoing -6.00 D and -7.00 D corrections were within ±1.00 D of intended refraction. Anterior stromal "haze," maximal at 6 months and diminishing thereafter, was detected in 110 patients (92%). However, best-corrected Snellen visual acuity was reduced in only 22 (18%). Good correlation existed between haze and regression (r = 0.68). Night "halos," due to the relatively small 4-mm diameter ablation zone, were reported by 94 patients (78%) in the early postoperative period, and, at 1 year, 12 (10%) declined treatment of the other eye because of persistence of this problem. Conclusion: Marked individual variation was found after photorefractive keratectomy. However, in low myopia, predictability is "fair," and the procedure is safe. The authors conclude that excimer laser photorefractive keratectomy holds considerable promise for refractive surgery in the future. Ophthalmology 1992;99: 1209-1219
The potential of the argon fluoride excimer laser in ophthalmic surgery is based on two characteristic features. First, operating at a wavelength of 193 nm, corneal tissue
Originally received: January 22, 1992. Revision accepted: March 16, 1992. From the Department of Ophthalmology, St Thomas' Hospital, London. Presented at the American Academy of Ophthalmology Annual Meeting, Anaheim, October 1991. Supported by the Iris Fund for Prevention of Blindness, London, England, including a research fellowship sponsored by the Iris Fund for Prevention of Blindness awarded to Dr. Gartry. Professor Marshall is a consultant for Summit Technology. Drs. Gartry and Kerr Muir have no proprietary interest in the subject matter of this article. Reprint requests to David Gartry, MD, FRCS, FCOphth, Moorfields Eye Hospital, City Road, London EC I V 2PD, England.
can be removed with a high degree of precision with virtually no damage to adjacent unexposed areas. 1-3 Second, the beam has a relatively large diameter (5 to 6 mm) and can be configured in any cross-sectional shape. When this beam passes through a computer-controlled expanding iris diaphragm, the corneal surface can be directly flattened to effect a reduction in myopia. This no-touch procedure has been termed photorefractive keratectomy (PRK).4 Computer-generated algorithms relating the diameter and depth of ablation needed for a given refractive change were devised based on the average amount of tissue removed per pulse. 5 Using these algorithms, a -7.00-diopter (D) correction can be achieved with removal of only 44 ~m of tissue (less than 10% of the axial corneal thickness) when a 4-mm diameter beam is used. The acute change and the healing response were studied in animals using experimental systems,6-1 3 and several ophthalmic excimer lasers are currently undergoing clinical evaluation. These
1209
Ophthalmology
Volume 99, Number 8, August 1992 Table 1. Patient Selection Criteria
systems were first used to treat superficial corneal pathology and the high degree of precision of tissue removal was confirmed. 14 More recently, treatment of refractive error, principally myopia, has been addressed. 15 - 23 This report presents the results of the first sighted eye cohort to undergo PRK in the United Kingdom. The study, which began in November 1989 at St. Thomas' Hospital, London, comprised 120 eyes (120 patients) with followup from 12 to 22 months (mean, 18 months).
Astigmatism Visual acuity
Materials and Methods
Age Contact lens history
Parameter Refraction
Laser A Summit Technology UV200 Excimed laser (Waltham, MA) with an emission wavelength of 193 nm, a fixed pulse repetition rate of 10Hz, and a radiant exposure of 180 mJ/cm 2 at the cornea was used. Early laboratory studies indicated that at this energy level approximately 0.22 !lm of corneal tissue was removed per pulse. To facilitate comparison among patients, this standard value was programmed in all procedures (although lasers manufactured since the beginning of the current study [November 1989] have incorporated a fixed value of 0.25 !lm/pulse in the algorithm). In addition, a standard ablation zone diameter of 4 mm (the maximum available at the beginning of the study) was selected in each case. There was no effluent removal system and no eye stabilizing system. The patient was required to fixate a target light within the laser aperture for the duration of the procedure (14 seconds on average). Human Studies After 2 years of experience in the treatment of corneal pathology, 14 Ethical Committee approval was given for a two-phase study; first, in blind eyes to allow assessment of the safety of the procedure and consolidate the surgical technique; and second, given satisfactory results, in sighted eyes to determine predictability and stability of the induced refractive change. The Blind Eye Study and preliminary sighted eye results have been reported in detail elsewhere. 15 Patient Selection Criteria Selection criteria are summarized in Table 1. Prospective patients were sent background information, in which the research nature of the project was emphasized. After questionnaire and interview stages, suitable patients underwent a detailed preoperative examination (Table 2). Study Design Depending on preoperative refraction, patients were allocated to I of 6 treatment groups (-2.00 D through to -7.00 D in whole diopter steps) with 20 patients per group. Within a group, all patients had an identical PRK treat-
1210
Ophthalmic history Occupation
Motivation
Informed consent
Criterion Patients with myopia (all degrees) with stable refraction; anisometropic patients included Less than 1.00 diopter correction Best-corrected visual acuity 20/30 or greater in both eyes Lower limit 24 years Contact lens wearers preferred due to likelihood of induced anisometropia; mainly soft lens wearers; patients with evidence of fluctuating refraction due to corneal distortion were excluded Patients with a history of ocular disease were excluded Individuals with exacting vocational visual requirements were excluded (e.g., pilots, professional drivers) Prospective patients were asked to state their reasons for wishing to be included in the trial; only those with realistic expectations were included Only those patients demonstrating a full understanding of the research nature of the trial were included; a detailed informed consent was signed prior to the surgery
ment (i.e., the same diopter correction was attempted). This allowed assessment of healing of identical ablation depths within each group. The range of preoperative myopia was -1.50 D to -17.50 D (spherical equivalent). Because the maximum treatment programmed was -7.00 D, and approximately one fifth of the cohort had myopia in excess of this, only partial correction was attempted in these patients. Anisometropic patients also were included and therefore emmetropia was not the aim in every case. To evaluate individual variation, all variables within our control were kept constant among patients: ablation rate (0.22 !lm/pulse), diameter of ablation zone (4 mm), surgical technique and postoperative treatment regimen (high-dose topical corticosteroids for 3 months; see below). Postoperative Management and Monitoring Immediately after the procedure was completed, slit-lamp examination was performed and the cornea was photographed (Fig 1). Oral analgesics (dextropropoxyphene/paracetamol [Distalgesic] X 2) were given immediately, and 4 tablets were supplied for home use. Topical cyclopentolate 1%,
Gartry et al . Excimer Laser Photorefractive Keratectomy Table 2. Schedule of Investigations and Follow-up Preoperative 1/7 5/7 2/52 4/52 6/52 8/52 10/52 3/12 4/12 6/12 9/12 1yr 18/12 Investigation Refraction Retinoscopy Autorefraction Keratometry Photography Halo measurement Endothelial specular microscopy Slit-lamp lOP Contrast sensitivity Videoimaging lOP
=
+ + + + +
+.... + + + +
+ +
+ +
+ +
+ +
+ +
+ +
+ + +
+ + +
+ + +
+ + +
+ + +
+ + +
+
+
+ +
+
+
+
+ +
+ +
+ + + + + + + + + + +
+ +
+ +
+ +
+ +
+ +
+ + +
+ + +
+ + +
+ + +
+ + +
+
+
+
+ + +
+
+ +
+
+ +
intraocular pressure.
phenylephrine 10%, homatropine 2%, and betamethasone with neomycin ointment were instilled, and the eyelids were taped and padded for 16 hours. The accepted role of topical corticosteroids in treatment of corneal disease, their capacity to modulate experimental wound healing (mainly linear incisions in rabbits), and limited experimental evidence that their use also might minimize corneal scarring after PRK9 resulted in Ethical Committee constraints such that a fixed regimen of topical dexamethasone 0.1 % 5 times per day for 2 months, tapering over the third month, was used in all patients (irrespective of refractive status or corneal clarity). This standardized regimen also ensured that comparisons could be made
Figure 1. Corneal appearance immediately after a -6.00 0 PRK procedure. The concentric ring pattern provides evidence of steady fixation and also can be used to assess the centration of the ablated zone in relation to the optical axis. The irregular light reflex beyond the ablation zone is the edge of the area of epithelial debridement.
among patients. (As a result of the current study, however, permission was granted for a randomized, double-masked trial with and without topical corticosteroids. The results of this trial will be reported soon). Topical chloramphenicol 0.5%, 4 times per day, also was used for the first 2 weeks. The follow-up intervals and assessment procedures at each stage are summarized in Table 2. During subjective refraction, reliance was placed on fogging techniques, in particular the + 1.00 D blur test, to ensure that an accurate end point was reached consistently. This test assumes that when a further, unnecessary + 1.00 D is added, the patient will be "fogged" to approximately 20/60. If, with this extra + 1.00 D, Snellen acuity is significantly better than 20/ 60, then too much minus has been prescribed and the spherical component can be revised accordingly. It was not possible to use cycloplegic refraction because of unwanted aberration effects from the paracentral cornea. Corneal clarity was subjectively graded on a scale from 0 to 4, as described in detail previously. IS Night halo effects, which were found to be due to the relatively small diameter of the ablation zone (4 mm), could be eliminated either by miosis or by interposing an appropriate negative lens. By using the internally illuminated standard Snellen visual acuity chart against a dark background, a subjective assessment of the magnitude of the halo effect could be made. After accurate refraction at each postoperative visit, patients were asked to estimate the width of the halo in comparison to the width of the Snellen chart. Increasing powers of negative lenses were then placed in the trial frame until the halo was eliminated. The minimum power of minus lens required to eliminate the halo was recorded. To determine its effect on the magnitude of the halo, an estimate of pupil size for each patient was made using a hand-held comparator (under the same clinical conditions used for measurement of the halo with the Snellen chart).
1211
Ophthalmology
Volume 99, Number 8, August 1992
Results Approximately 2 hours after the procedure, all patients reported pain r~n~ng from mild discomfort to moderately sev~re. The maJonty slept well, although a significant minonty woke on one or more occasions. Most used all 6 Di~talgesic tablets supplemented with proprietary analgesICS. The pain settled within 24 hours and all patients reported that the eye felt normal after 4 days. By the fifth day, re-epithelialization was complete. Unaided vision was reported to be considerably improved by day 4, and by week 1 accurate subjective refraction was possible. Best spectacle-corrected visual acuity returned to preoperative levels by week 2. On direct questioning, no patient reported diurnal variation. Postoperative Refraction A series of graphs relating mean change in refraction to follow-up time is shown in Figures 2A to F. All patients had. a significant decrease in myopia in the first postoperatIve week. A mean overcorrection was noted propor~ional to the magnitude of the attempted correction (e.g., In the -3.00 0 treatment group the mean change was 5.500, while in the -7.00 D group it was 10.25 D). This overcorrection reduced progressively between the second and twelfth postoperative weeks. Further regression occurred in some individuals in all groups but was more common in the -6.00 D and -7.00 0 groups (Figs 2E and F). In the -7.00 D group, the degree of immediate postoperative change also was related to the amount of preoperative myopia. In analyzing data, patients in this group were subdivided into those whose preoperative refractive error was less than -10.00 D (mean, -7.600; n 10) and those whose preoperative refractive error was greater than -10.00 D (mean, -12.94 D; n = 10) (Fig 3). The mean change at the first postoperative visit for the former group was 9.00 D (an overcorrection of 2.00 D), and for the latter group 12.25 D (an overcorrection of 5.25 D). The -2.00 D, -3.00 D, -4.000, and -5.00 D groups demonstrated minimal change in mean refraction after the third month (Figs 2A to D). In contrast, in the -7.00 D group, a more marked regression toward the preoperative level occurred throughout the first 6 months after which mean refraction stabilized (Fig 2F). Although the mean change in refraction achieved at 3 months in all groups was close to that intended, there was considerable individual variation (Fig 4). The refraction results at 18 months are summarized in Table 3 and the unaided visual acuity, in those patients in whom a full correction was attempted, are summarized in Table 4. There was a ~igh direct correlation between keratometry and refractIon. No statistically significant change in astigmatism was fo~nd when compared with the preoperative amount (twot~Iled Student's t test, P> 0.05). Retinoscopy was relatIvely easy to perform postoperatively, and a well-defined central zone of altered reflex was clearly visible. The central "with" movement was surrounded by an annulus of "against" movement. Retinoscopy was most useful in the
1212
Table 3. Accuracy of Photorefractive Keratectomy (Follow-up 1 Year) Patients within Stated Tolerances
Treatment Group (D)
±O.50D
±1.00D
Patients Achieving >1/2 Attempted Correction (%)
-2.00 -3.00 -4.00 -5.00 -6.00 -7.00
75% 45% 20% 30% 25% 5%
95% 70% 40% 50% 40% 20%
95% 90% 65% 80% 60% 70%
early postoperative period to confirm the amount of overcorrection present, and to minimize the tendency to accommodate before subjective refraction. Visual Acuity and Anterior Stromal Haze Correlations At 1 year, 98 patients (82%) had best spectacle-corrected visual acuity equal to or better than preoperative visual acuity. Eighteen patients (15%) lost between liz and 1 line of Snellen acuity, while 4 patients (3%) lost 2 lines (Table 5). These 4 patients had undergone either -6.00 0 or -7.00 ~ procedures (ablation depths of 39 and 44 JLm, ~esp~ctIvely). As expected, the irregularities of healing epIthelIum caused a slight reduction in clarity in the first postoperative week. After this, the cornea was clear until the 3- to 6-week stage, when a faint, diffuse subepithelial stromal haze appeared that increased to a maximum by 6 months. This change in corneal clarity (haze) was detected in 110 patients (92%) by the third month. The magnitude and time course of this haze, as determined by our subjective grading method, are shown in Figure 5. There was only limited statistical correlation of haze with age (r = -0.32) and poor correlation with ablation depth (r = 0.22) despite the clinical impression that older patients demonstrated less haze, and higher order corrections (deeper ablations) resulted in more marked haze. There was good correlation, however, between haze and the regression of refraction (r values for the-2.00 D to -7.00 D groups = 0.49, 0.75, 0.83, 0.36, 0.86, and 0.76, respectively; mean correlation, 0.68). In addition in those patients in whom correction should have prod~ced emm.etropi~ (n = 66), unaided vision was highly correlated WIth reSIdual ametropia at 1 year (r = 0.94). Side Effects ~able 5 identifies complications and their percentage inCIdence. The one case of significant decentration of the ablation zone occurred in the third patient to be treated. Fixation was maintained, but viewing angle altered when the patient moved steadily away from the surgeon. This could have been avoided by interrupting the treatment
Gartry et al . Excimer Laser Photorefractive Keratectomy Change in Refraction following -2.00D PRK (Ablation depth
Change in Refraction following -3.000 PRK
= 17um) 14
8 mean preoperative refraction
Ci
r::
0
-2.10D
6
(n
= 20)
«J
.= Q)
CI
2
r::
o
10
-
mean preoperative refraction
.........
--~
n = 20
t1
.&:.
. ...
0 0
8
Q)
a:
(J
A
-
-3.86D
«J
4
«J
12
r::
o
0
.........Q) a:
e
(Ablation depth = 23 urn) ~----_.-----,----_r----_r----_r----_..
3
6
.=
r r
1 9
12
f
15
6
Q)
4
CI
r::
.«J
.&:.
2
(J
o
18
B
Months
3
6
9
12
15
18
Months Change in Refraction following -5.000 PRK
Change in Refraction following -4.000 PRK (Ablation depth
o
= 28um)
(Ablation depth = 33um)
14
12 mean preoperative refra c tion
Cl
r:: 0 .;:;
12
mean preoperative refraction -4.00D
10
(n = 20)
-a:
8
r:: Q)
CI
r::
.=
6
Q)
4
(J
.&:.
2
(J
0 0
6
9
12
15
0
D
r-r----,-----,----_.-----,----_r----_r~
-7 .00D
10
(n= 20)
.&:.
12
15
18
mean preoperative refracti on Cl
r::
.2
0«J
1\
8
-
6
------------------------
-a:....
12
-10 .26D
(n = 20 )
10 8
Q)
.=
6
CI
4
Q)
4
r::
r::
(J
9
(Ablation depth = 44um)
14
CI «J
6
0
Change in refraction following -7.000 PRK
c: Q)
2
Change in Refraction following -6.000 PRK
«J ....
Gi a:
---------------r------
f--4
Months
r::
.g
6
Months
mean preoperative refraction Cl
8
0
18
(Ablation depth = 39um) 12
= 20)
r::
«J
C
(n
CI
«J
.&:.
0
.t: Q) a:
Q)
-7 .13D
10
r::
0«J
0
«J ....
Cl
«J
.&:.
2
(J
0
E
0
3
6
9
Months
12
15
18
2
0
F
0
3
6
9
12
15
18
Months
Figure 2A to F. Graphs relating mean change in refraction to time. Initial overcorrection in all cases is followed by regression to a plateau at between 3 and 6 months. The error bars are ±1 standard deviation. (Minimum follow-up for all patients is 12 months; maximum, 22 months; mean, 18 months).
1213
Ophthalmology
Volume 99, Number 8, August 1992
until the patient was repositioned. This patient experienced a greater amount of halo effect (vide infra) below lights, which gave rise to difficulty driving at night. Fourteen patients (12%) had a measurable rise in intraocular pressure (lOP) detected at the 4- to 6-week stage (topical corticosteroid responders). Of these, 3 reported transient loss of vision on rising rapidly from a chair and were found to have lOP in excess of 45 mmHg. Topical corticosteroids were discontinued, and lOP returned to normal within 2 weeks. Those with lOP in excess of 30 mmHg were treated with topical timolol maleate 0.25% twice a day for 1 month.
Individual Variation following PRK ( n = 120 at one year foll o w-up)
e z
o i= u W
~
o U o >
W
Desired
7
,........ ........ ..-- Correction 0 ....... / /
6 o
8
5
e..,.................
4
...... o
;"//
0...... . . . .
/"'"0
~
....... - B
Mean
0
g
iii
-4
-5
...--
Correction
Achieved
3
w
J:
u
Halo Effect (Low Illumination Positive Spherical Aberration) On direct questioning, 94 patients (78%) reported halos around sources oflight at night in the early postoperative period. The magnitude of this halo effect was highly directly correlated with pupil size (r = 0.79) and the amount of the induced refractive change (r = 0 .71) (Figs 6A and B). Those patients with greater overcorrection in the early postoperative phase experienced the most marked halo effect. With regression in refraction in the first 3 months subjective appraisal of the halo effect and a reduction in the power of the negative lens required to eliminate it indicated that the magnitude of the halo became less with time. However, at 1 year, 12 patients (10%) found that the effect persisted and was severe enough to interfere with night driving. They therefore declined to have the other eye treated. The effect was found to be unrelated to the lllagnitude and time course of the anterior stromal haze.
Change in Refraction following -7.00D PRK 14
0
'"
C>
0
4
2
C> Gl
Gl
Z
60 GlQ.'I;j
