Laser thermal keratoplasty for the treatment of photorefractive keratectomy overcorrections

Laser Thermal Keratoplasty for the Treatment of Photorefractive Keratectomy Overcorrections

Mihai Pop, MD Objective: To evaluate the results of holmium:YAG laser thermal keratoplasty (LTK) treatment for overcorrection of myopia after a photorefractive keratectomy (PRK) treatment. Participants: Thirty-six eyes (33 patients) were treated with a nontouch holmium:YAG laser (Sunrise Technologies, Model LTK, Freemont, CA) because of hyperopia (mean t standard deviation of +2.06 diopter [D] -+ 0.75, ranging from +l .O to +3.5 D) following a PRK treatment. A control LTK group treated for primary hyperopia, who had preoperative refraction values not statistically different from the PRK + LTK group, was used for comparison. The number of spots applied varied from 8 to 24, and the energy used was 200 to 240 mJ. A Intervention: maximum of three rings of four to eight spots were placed between 6 and 8 mm from the visual axis. Results: Twelve months after the LTK retreatment for PRK patients, mean refraction was +1,14 D ? 1.09. Regression from 1 to 12 months was 0.5 D ? 1.1. At 12 months, 50% of eyes were within 1 D of emmetropia; 93% of eyes had uncorrected visual acuity (UCVA) of 20/40 or better; and 24% of eyes had UCVA of 20/20 or better. Refraction was not stable for 11 eyes (34%) that regained original sphere values or higher. Best-corrected visual acuity was not affected, and haze was not increased nor decreased by the procedure. Conclusions: Twelve months after an LTK retreatment for an initial PRK, two thirds of the retreated eyes did not need further retreatments. However, clinical data showed that LTK should be kept for +l to +2 D of hyperopia for PRK overcorrection retreatments. Ophthalmology 1998; 705:926-937

Photorefractive keratectomy (PRK) has become a common method to correct myopia. While most eyes are corrected to near emmetropia, a certain number end up with undercorrection due principally to regression, haze, or both.‘,2 However, a small percentage of eyes are overcorrected. The purpose of this study is to evaluate the use of holmium:YAG laser thermal keratoplasty (LTK) when retreating PRK-overcorrected myopic eyes. To evaluate this possibility, LTK was proposed to patients who were overcorrected by at least 1 diopter (D). Koch et al’,4 reported safe, effective, and persistent correction of low hyperopia (+ 1 D to +3, or +5 D) in the majority of LTK-treated eyes for primary hyperopia. Other authors reported various regression effects using LTK technology for primary hyperopia5,” or for treatment of astigmatism.7m9 Nevertheless, LTK has been classified Originally Revtston From

received: December 2, 1996. accepted: October 17, 1997.

the Michel

Pop Clinics,

Montreal

and Hull,

Canada.

Presented in part at the Annual Meeting of the Amerrcan Ophthalmology, Chicago, October, 1996. The author this study.

has no proprietary

interest

Academy

in any of the materials

used m

Reprint requests to Mihai Pop, MD, Michel Pop Climes, 9001 de l’Acadte Blvd. N. Surte 102, Montreal, PQ, Canada H4N 3H.5.

926

of

as safe and effective for low hyperopia in the majority of eyes4 One of the hypotheses underlying this study is that the disappearance of the Bowman’s layer after initial PRK could have an impact on the amount of correction by an LTK and its stability. The LTK is based on the production of a series of intrastromal coagulation points. So, it was hypothesized that the Bowman’s layer might play a role in the ultimate partial relaxation of these points with time. Since the Bowman’s layer is removed during a PRK treatment, it was also hypothesized that LTK treatment could be more stable for PRK overcorrections than for primary hyperopia. This hypothesis could be verified when comparing regression of the natural LTK hyperopic group with the PRK-treated/LTK-retreated eyes (PRK + LTK). Another hypothesis, contrary to the first one, is that LTK stability for PRK overcorrections would implicate the thinning of the cornea. Since PRK removes subepithelial stroma, the cornea is actually thinned by the PRK procedure. The stability of the coagulation points performed by LTK treatment might be influenced by the remaining stroma, since the strength of these points depends solely on amount of stroma being coagulated. Therefore, LTK points for PRK overcorrections would lose their strength and sharpness over time. LTK for natural hyperopia would be more stable than LTK for PRK

Pop * Laser Thermal

Keratoplasty

for PRK Overcorrection

Table 1. Percentage of Eyes with Specific Best Spectacle-corrected Snellen Visual Acuity (BSCVA) before and after LTK PRK

+ LTK

LTK

20120 or Better (96)

2Ol25 or Better (o/o)

20140 or Better (%I

45 57 61 61 75 71

79 82 83 85 84 87

88 97 100 96 94 94

PRK = photorefractwc keratectomy; treated with LTK; N = no. of eyes.

LTK

= laser thermal

keratoplasty;

overcorrections. This hypothesis could also be verified by comparing a natural LTK hyperopic group with the PRK + LTK group.

Patients and Methods Laser thermokeratoplasty, using a nontouch holmium:YAG laser (Sunrise Technologies, Freemont, CA), was proposed to and accepted by 33 patients (36 eyes) who had become hyperopes after myopic PRK treatment. The LTK post-PRK group was compared with a natural hyperopia population of 40 eyes (27 patients) with a comparable level of refraction error. This population had no PRK treatment prior to the LTK. Preoperatively, manifest refraction, best spectacle-corrected Snellen visual acuity (BSCVA), cornea1 topography, and slitlamp clinical examination of the eye and retina were performed. The experimental nature of the procedure was explained and discussed with the patients. If they agreed to participate, they were required to sign a consent form describing the procedure, the possible outcomes, and the side effects. All eyes had normal anterior chambers and lenses prior to LTK. The Sunrise Technologies holmium solid-state laser delivers infrared light pulses of 2.1 pm in wavelength at a frequency of 5 Hz through a fiberoptic system that can be set to produce one to eight spots simultaneously. The number of pulses, the energy densities, and the number of spots were applied according to the level of correction. To correct hyperopia, four to eight spots per ring, up to three rings, were applied in concentric circles with a diameter ranging from 6 mm to S mm from the visual axis. The spots could be applied in two ways: (1) in a radial pattern and (2) in a skewed-staggered pattern, with a rotation angle of 22.5” between rings. When astigmatism was present, four coagulation spots along the flattest meridian were applied. The number of pulses varied from five to ten. Total number of spots varied from 8 to a maximum of 24. The pulse energy varied from 200 to 240 mJ. Eyes were anesthetized with topical drops of 0.5% proparaCaine (Alcaine, Alcon, Montreal, Canada). A lid speculum was used. The cornea was left to dry for 3 minutes. Then, the cornea was further dried with a microsponge. Self-fixation was the rule in most of the cases. The location of the spots was determined by a series of eight helium-neon target lights that were delivered through the same fiberoptic system. The duration of each treatment with a maximum of eight spots took approximately 2 seconds.

N

20120 01 Bate? (1%)

20125 01 Better (9,)

20140 or Better (S)

N

33 33 23 26 32 31

48 55 61 60 63 62

74 76 78 80 84 86

95 97 100 97 97 100

39 38 28 35 32 29

PRK

+ LTK

= PRK overcortectmx

treated

wth

LTK;

LTK

= hyperopm

Postoperative care included drops of ofloxacin 0.3% (Ocuflox, Allergan, Montreal, Canada) and ketorolac tromethamine 0.5% (Acular, Allergan) for 3 days. Follow-up examinations were performed after 72 hours, then at I, 2, 3, 6, and 12 months. The examinations included subjective manifest refraction, Snellen best spectacle-corrected and uncorrected visual acuity (BSCVA and UCVA), intraocular pressure, and cornea1 topography. Haze was assessed using a 0 to 4 scale (clear to completely obscured), as proposed by several authors.‘“,” Evaluation of haze did not include the residual LTK points on the cornea that could be seen following treatment. Some eyes were retreated for a second time with the holmium laser between 6 and 12 months postoperation, when no refractive change was seen following the first LTK treatment. Some of these post-LTK eyes were also retreated using the Nidek EC-5000 hyperopia module (Nidek Co., Tokyo, Japan). The results were analyzed using Student’s t test, paired t test, and correlation coefficient with the Simstat Version 3.5 statistical program (Provalis Research, Montreal, Canada). Foxprow 2.6 software was used for compilation of data. Vector analysis of astigmatism was done with the software RefractTools, Version 1.38 (provided by Julian Stevens, MD, Moorfields Eye Hospital).

Results Mean age +- SD of the PRK + LTK group was 39.4 years t8.0, ranging from 19 to 54 years. Forty-five percent of the patients were female and 55% were male. Mean refraction ? SD before treatment was +2.07 D -C 0.75, ranging from + I to +3.50 D. Astigmatism varied between 0 and -3.5, with a mean + SD of -0.78 D 2 0.95. Eleven eyes (30.5%) were retreated after the study for further enhancement. Before LTK, 29 (88%) eyes in the PRK + LTK group had 20/40 or better BSCVA. The remaining four eyes that initially had 20/40 lost 1 line after the PRK procedure and started at 20/50 before LTK. Three of these eyes had refraction values of - 10 D or greater before PRK. After PRK, one of these eyes had a central island prior to LTK. Mean age ? SD of the natural hyperopic group (LTK only group) was 44 years ? 11, with the same percentage of males (55%). Mean refraction 2 SD before treatment was +2.32 D t 0.70. The two groups were not statistically significantly different (P = 0.14) in terms of mean preoperative refraction values. Two eyes had amblyopia and had BSCVA values below

927

Ophthalmology

Volume 105, Number 5, May 1998

Table 2. Percentage of Eyes with Specific Uncorrected PRK

1 2 3 6 .2

LTK

20/25 or Better (%)

20140 or Better (%)

19 30 25 28 24

54 43 50

92 91 92 93 93

PRK = photorefractive keratectomy; treated with LTK; N = no. of eyes.

LTK

= laser thermal

keratoplasty;

Table 3. Number of Eyes Having Lost 1 and 2 Lines of Best Spectacle-corrected Snellen Visual Acuity after Laser Thermal Keratoplasty Treatment PRK + LTK

LTK

2 Lines

1 Line

N

2 Lines

3 1 0 0

10 5 2 4 2 1

33 31 21 26 30 30

2 2 1 1 1

i

1 Lme 2 1 4 4 2

N 40 37 27 34 31 28

PRK = photorefractive keratectomy; LTK = laser thermal keratoplasty; PRK + LTK = PRK overcorrectlons treated with LTK; LTK = hyperopia treated with LTK; N = no. of eyes; - = no loss before initial LTK.

928

20125 or

N

20/20 or Better (%I

26 23 24 29 29

22 17 10 17 14

PRK + LTK

20/40 (20/80) before LTK. Astigmatism varied from 0 to -6.5 D, with a mean cylinder t SD of - 1.48 D t 1.87. Fifteen eyes (37.5%) were retreated after the study for further enhancement. Best spectacle-corrected visual acuity was not affected by the procedure. Percentages of patients within BSCVA 20140 or better and 2012.5 or better were 88% and 79%, respectively, prior to LTK, while percentages were 94% and 87% 12 months after retreatment (Table 1). For the natural hyperopic LTK group, percentages of patients within BSCVA 20140 or better and 20125 or better were 95% and 74%, respectively, before treatment, while percentages were 100% and 86% 12 months after treatment (Table 1). Significant correlations were found for BSCVA values before and after LTK treatment for the natural hyperopia LTK group (P < 0.01) and PRK + LTK group (P < 0.001) at 12 months postoperative, meaning that BSCVAs were statistically the same before and after LTK (no Logmar scale was used for these correlations). Table 2 shows the LJCVA post-LTK. After 1 year, in the PRK + LTK group, 93% of eyes had 20140 or better UCVA and 24% of eyes had 20120 or better UCVA. At 12 months postoperation, no eyes lost 2 lines of BSCVA for the PRK + LTK group, and one eye lost 2 lines of BSCVA for the LTK group (Table 3). The loss of 1 line was below 2 eyes for both groups. Interestingly, the number of eyes having lost 1 and 2 BSCVA lines decreased after LTK retreatment for the PRK-treated eyes. However, no data support the fact that this could be due to LTK itself rather than the healing time following PRK surgery. The very low rate of eyes having lost 1 or 2 BSCVA lines reflected the fact that the LTK procedure was almost free of complications on visual acuity.

Before 1 mo 2 mos 3 mos 6 mos 12 mos

before and after LTK

+ LTK

20/20 or Better 6) mo mos ma mos mos

Snellen Visual Acuity (UCVA)

= PRK overcorrectlons

(W

20/40 or Better (%)

N

35 46 35 31 39

81 79 81 62 71

32 24 31 29 28

Better

treated

with

LTK;

LTK

= hyperopta

A scattergram representing the results of the sphere at 6 months compared to its value before LTK shows a different response between the two groups (Fig 1). In the group in which LTK was done on natural hyperopic eyes, only three eyes (9%) at 6 months and one eye (3%) at 12 months showed no improvement in their refractions (Table 4). In the eyes treated for PRK overcorrection, 11 eyes (34%) at 6 months and 11 eyes (35%) at 12 months did not gain any benefit from the LTK procedure (Table 4). At the 12-month follow-up examination, the differences in refraction values in the PRK + LTK group (mean +- SD of + 1.14 D + 1.09) and the LTK only group (mean + SD of + 1.52 D -t 0.79) were statistically significant (P < 0.01) when compared with their preoperation refraction values (Table 5). The mean refraction values at 12 months were slightly higher in the LTK group than in the PRK + LTK group. However, at 12 months’ follow-up, both groups were not statistically different (P > 0.12). Regression was calculated on an individual basis. Refraction value differences from 1 to 12 months for each eye were calculated. Mean individual 1 to 12 months regression ? SD was 0.47 D -t 1.13 for the PRK + LTK group and 1.13 D 2 0.92 for the LTK group. A Student’s t test showed that these regression factors were not statistically equal (P = 0.02), with the LTK natural hyperopia group having the highest regression. In the PRK + LTK group, no multiple correlation was found to explain the relation between refraction values obtained at 12 months using refraction values before LTK treatment and before PRK and LTK treatments. Table 6 shows the different percentages of eyes that were at kO.5 D and ? 1 D of emmetropia for the two groups throughout the follow-up. During the first month, a higher percentage of eyes in the LTK group (42%) than the PRK + LTK group (33%) was within 20.5 D of emmetropia. But at 12 months, a higher percentage of eyes in the PRK + LTK group (25% and 50%) than the LTK group (17% and 38%) was within 20.5 D and t 1 D of emmetropia, respectively. When astigmatism was present, the mean refraction value ? SD changed from -0.78 D LO.95 before surgery to -0.85 D kO.64 at 12 months in the PRK + LTK group, but this change was not statistically significant (P = 0.70, using t test with unequal variances). In the LTK group, the reduction from - 1.48 D -C 1.87 to - 1.34 D + 1.48 was also not statistically significant (P = 0.73). A vector analysis, made at 12 months, showed astigmatism undercorrection in both groups. The mean difference vector magnitudes were -0.86 D and -1.35 D in the PRK + LTK and in the LTK groups, respectively. The mean magnitudes of cylindrical error were -0.21 D and -0.97 D, respectively. Mean Alpins’ success indexes were 0.99 and 0.67,

POP*

Laser Thermal

Keratoplasty

for PRK Overcorrection

B -

5

N = 29 4 F 3 .r( a2 22 1 -2 -3 I

8 II

1

0

2 Intended

2

1

3

3

4

Intended

versus 12 months postoperative showing achieved versus mtcnded Figure 1. A, scattergram of the sphere values (dioptcrs [I)]) h e f ore LTK retreatment correction toward emmetropia for the PRK + LTK group. B, scattergram of the sphere values (LI) before LTK treatment versus 12 months postopcratlvr showing achlcved versus intended correctmn toward emmctropia for the LTK group. Straight lines indicate levels wlthm ‘0.5 D of the mtencicd correction. Number of eyes (n) is shown.

respectively, for the PRK + LTK and LTK groups.Therefore, usingthese results, we cannot conclude that astigmatism was corrected adequately in the majority

Mean haze values for the

of eyes.

+ LTK group were not statistically different from levels before LTK and levels 6 and 12 months after treatment (P = 0.30 and P = 0.30). Mean haze values were always below 0.5 (Table 7). Generally, haze was not a complication. No haze was seen in the natural hyperopic LTK group, except for LTK treatment points on the cornea that were not considered as scarification. LTK points that could be seen on the cornea after the treatment also seem to fade with time, although no data from this study are available on this since it was not methodically evaluated from the beginning. No statistical tests showed significant correlations between age of the patient and refractive outcome at 6 and 12 months postoperative for both the PRK + LTK and LTK groups. Since in this study hyperopia ranged from + 1 .O to +3.5 D, this range may not be large enough to test the correlation of hyperopia with age. PRK

However, not all eyes need a retreatment becauseof the accommodation factor. One study reported that retreatments f’or undercorrections had resulted in acceptable final refractions for at least 50% of the cases that were missed in the first instance.” In another study on myopia retreatments with the Visx 20/20B excimer laser (Visx Co., Santa Clara, CA), 84% of eyes had acceptable final refraction values when regression did not imply a high formation of haze.12That study reported that overcorrections of 1.25 to 2.00 D could account for 8% of eyes. To avoid a high percentage of overcorrection, one author proposed reducing the amount of correction, even if it means a higher percentage of undercorrections.” The risk of retreating overcorrections would still be present, even if the inadequately healing population could be hypothetically isolated in the follow-up results of the first PRK.

Discussion Among all of the complications from myopic PRK corrections, over- and undercorrections generally rank first.

Table

5. Mean

Sphere

Values

before

PRK

+ LTK

Mean

Table 4. Number of Eyes Having Sphere Values Equal or Greater

than

before PRK

6 mos 12 mos

LTK

Treatment

after

+ LTK

LTK

Eyi3

N

Eye,

N

11 (34%) 11 (34%)

32 32

3 (9%) 1 (3%)

32 29

PRK = photorefractive keratectomy; LTK = Laser thermal kcratoplasty; PRK + LTK = PRK overcorrectmns treated with LTK; LTK = hyperopia treatetl with LTK; Eyes = no. of eyes havmg sphere equal or greater than preoperative sphere; N = no. of eyes m the study.

Before 1 mc 2 mob 3 mob 6 mos 12 mob No. of rctreatments post LTK

t2.06 to.68 +0.96 to.95 +1.24 t1.14

Mean N

t +k t rt t

0.75 0.88 0.97 1.10 1.08 1.09

11 (30.5%)

LTK

Treatment LTK

?z SD 0)

6 or 12 Months

and after

36 33 24 26 32 32 36

2 SD N

0) t2.31 +0.49 + 1.06 +1.19 +1.38 t1.52

z k + f t t

0.70 0.99 0.90 0.77 0.85 0.78

15 (37.5%)

40 38 29 35 32 29 40

PRK = photorefractlve kcratectomy; LTK = laser thermal kcratoplasty; PRK + LTK = PRK overcorrcctions treated with LTK; LTK = hypempia treated with LTK; SD = standard devlatlon; N = no. of eyes.

929

Volume 10.5, Number 5, May 1998

Ophthalmology Table

6. Percentage PRK

of Eyes wlthin

to.5

and

Diopters

of Emmetropia

+ LTK

LTK

20.5 D (%)

k1 D (%)

N

0 33 21 23 16 25

17 67 42 46 50 50

36 33 24 26 32 32

Before 1 mo 2 mm 3 mm 6 mos 12 mos

?l

r0.5 D (%)

-cl D (%)

N

31 25 18 17

10 71 44 54 41 38

40 38 29 35 32 29

PRK = photorefractive keratectomy; LTK = laser thermal keraroplasty; PRK + LTK = PRK overcorrectlons treated with LTK; LTK = hyperopn treated with LTK; N = no. of eyes.

Thermokeratoplasty was considered unpredictable and a potential cause of severe side effects before the advent of the holmium:YAG laser.14Koch et al6 first reported spherical equivalent changes from + 1.6 D before LTK to +0.7 D (a 0.9 D difference) at 6 months postoperative with a laser similar to the one used in this study. But regression after longer periods of time was a concern for these authors. The present study showed that for low hyperopia, a relative improvement of 1 D of mean manifest refraction could be seenafter 12 months. Individual responseto the treatment seemedto be very variable, up to 1 D of SD, and some regression seemedto result from the procedure, at least for the LTK group. As for the PRK -t LTK group, the mean improvement and regressionfactors were less than 1 D, although 50% of patients were within 1 D of emmetropia. These results show that the amounts of correction attainable with the algorithms used were not different between first LTK treatment and PRK + LTK retreatment if we look at the 6- and 12-month results. In both groups, an average of 1 D or less was gained. Thompson et al5 proposedLTK to correct astigmatism. They were able to correct up to 3.50 D with their technique. Using the same approach, we could not help subjects with residual astigmatism after a surface PRK. We cannot recommend LTK to correct astigmatism in view

Table

7. Mean

Haze Values and Number of Eyes Having Greater than 1 after LTK* PRK

Mean Haze Before 1 ,110 2 mos 3 mos 6 mos 12 mos

0.32 0.21 0.28 0.18 0.20 0.21

t 0.47 & 0.44 2 0.56 t 0.41 2 0.47 % 0.38

+ LTK

Haze

LTK

Eye3with Haze 21

N

Mean Haze

Eyes wxh Haze 21

N

: 5 3

36 33 24 26 32 32

0 0 0 0 0 0

0 0 0 0 0 0

40 38 29 35 32 29

i

PRK = photorefractive keratectomy; LTK = laser thermal keratoplasty; PRK + LTK = PRK overcorrections treated with LTK; LTK = hyperopia treated with LTK; N = no. of eyes. * Haze was assessedusing a 0 to 4 scale.

930

of our results. It is possible that the number of points used to treat astigmatism might not be enough, and subsequently, not strong enough to remain unimpaired. During the course of this study, two thirds of the eyes did respond to laser thermokeratoplasty, but some eyes had to be retreated using the same procedure or with another type of retreatment. A possible conclusion concerning these eyes could be that the effect of the coagulation points is not maintained over time and losesstrength. However, no effect was seen on BSCVA values or loss of BSCVA lines, keeping the procedure relatively safe in terms of side effects on visual acuity. Surface PRK changes the anterior portion of the cornea. The partial or complete disappearanceof the Bowman’s layer in the treatment zone has been implicated in the formation of scar tissue or cornea1haze. Some authors have speculated that it might have a structural importance for the integrity of the cornea1 shape.15Our hypothesis was that a partial removal of the Bowman’s layer or the thickness of the cornea following PRK might have an impact on LTK regression. The principle of LTK is the production of coagulation spots in the stroma. The regression seen in the months following the procedure might be partly due to the relaxation of the coagulated collagen strands. In the following months, these points could lose their sharpnessand become larger, more diffuse, or disappear. In view of the initial hypothesis made in this study, since LTK post-PRK seemedto provoke significantly less regression than LTK alone, the results might lead to the conclusion that the relaxation of coagulated collagen points could be due to the Bowman’s layer. The results would not favor the hypothesis that the thinning of the cornea due to PRK could influence LTK stability, since regression in PRK + LTK cases was lower than in the natural LTK group. Rather, the results would favor the hypothesis that collagen points would be strengthened by the thinning of the cornea. However, the improvement of refraction was low for both groups, reflecting the fact that the conclusion on the Bowman’s layer hypothesis is fragile and could be challenged. This is corroborated by the fact that more eyes in the PRK + LTK group than in the LTK group did not gain any benefit from the procedure. Overcorrected PRK patients are usually unhappy pa-

Pop * Laser Thermal

Keratoplasty

tients. The benefit for overcorrected PRK patients is an important issue to consider. Until a more effective treatment of hyperopia can be usedclinically, LTK is an effective method to correct small overcorrections due to PRK. However, in light of the results of this study, the use of LTK for PRK overcorrections should be kept for cases below +2 D. New algorithms or procedures are required to extend the use of the holmium laser for PRK overcorrection retreatments. The treatment of hyperopia and the retreatment of myopia overcorrections may also be improved by erodible masks or laser scanning technology with hyperopia modules. As studiesgo on, the best way to treat hyperopia and myopia overcorrections should reveal itself. Still, one of the best methodsis to avoid overcorrections, even if it meansa higher percentage of undercorrections. However, as noted by other authors and as shown in this study, the useof LTK technology is generally safe regarding the loss of BSCVA lines, with very few side effects and haze complications.5”6 Acknowledgments. The author thanks Julian Stevens, MD, FRCOphth, for the Refract Tools 1.38software,and Yves Payette, MSc, for his suggestions,contributions,and analysis of data.

References Durrie DS, LesherMP, CavanaughTB. Classificationof variableclinical responseafter photorefractivekeratectomy for myopia. J Refract Surg 1995;11:341-7. Durrie DS, SandersDR, SchumerDJ. Computerizedcornealtopographyof surfaceablationswith the EyeSysSystem. In: Salz JJ ed. McDonnell PJ, McDonald MB, assoc. eds. Cornea1Laser Surgery. Philadelphia:Mosby 1995; 109-29. Koch DD, Berry MJ, VassiliadisA, et al. NoncontactHolmium:YAG laser thermal keratoplasty. In: Salz JJ ed. McDonnell PJ, McDonald MB, assoc.eds.Cornea1Laser Surgery. Philadelphia:Mosby 1995;247-58.

for PRK Overcorrection 4. Koch DD, Abarca A, Villarreal R, et al. Hyperopiacorrec-

tion by noncontact Holmium:Yag laser thermal keratoplasty. Clinical study with two-year follow-up. Ophthalmology 1996;103:73l-40. 5. ThompsonVM, SeilerT, Durrie DS, CaranaughTB. Holmium:YAG laser thermokeratoplastyfor hyperopia and astigmatism:an overview. Refract Cornea1Surg 1993;9(2 Suppl):S134-7. 6. Tutton MK, Cherry PM. Holmium:YAG laserthermokeratoplastyto correcthyperopia:two yearsfollow-up. Ophthalmic Surg Lasers1996;27(5Suppl):S521-4. associated with myo7. Cherry PM. Treatmentof astigmatism pia or hyperopiawith the holmiumlaser:secondyear follow-up. OphthalmicSurg Lasers1996;27(5Suppl):S4938. 8. Cherry PM. Holmium:YAG laserto treat astigmatism asso-

ciatedwith myopia or hyperopia.J Refract Surg 1995;1l(3 Suppl):S349-57. 9. HennekesR. Holmium:YAG laser thermokeratoplastyfor correction of astigmatism.J Refract Surg 1995;1l(3 Suppl):S358-60. 10. FantesFE, HannaD, Waring GO III, et al. Wound healing after excimer laserkeratomileusis(photorefractivekeratectomy) in monkeys.Arch Ophthalmol 1990;108:665-75. 11. PopM, Aras M. Multizone/multipassphotorefractivekeratectomy: six month results. J Cataract Refract Surg 1995;21:633-43. 12. Pop M, Aras M. Photorefractivekeratectomyretreatments for regression. One-year follow-up. Ophthalmology 1996;103:1979-84. 13. Gimbel HV, Van WestbruggeJA, JohnsonWH, et al. Visual, refractive, and patient satisfactionresultsfollowing bilateral photorefractivekeratectomy for myopia. Refract Cornea1Surg 1993;9(2Suppl):SS- 10. 14. Aquavella JV, Smith RS, Shaw EL. Alterations in cornea1 morphologyfollowing thermokeratoplasty.Arch Ophthalmol 1976;94:2082-5. 15. Jain S, Azar DT. Extracellular matrix and growth factors in cornea1 wound healing. Curr Opin Ophthalmol 1994; 5:3- 12. 16. Ariyasu RG, Sand B, Menefee R, et al. Holmium laser

thermal keratoplastyof 10 poorly sightedeyes. J Refract Surg 1995;11:358-65.

931