Hyperopia Correction by Noncontact Holmium:YAG Laser Thermal Keratoplasty

Hyperopia Correction by Noncontact Holmium:YAG Laser Thermal Keratoplasty

Hyperopia Correction by Noncontact Holmium:YAG Laser Thermal Keratoplasty United States Phase IIA Clinical Study with a l . . year Follow . . up Dougl...

1MB Sizes 0 Downloads 4 Views

Hyperopia Correction by Noncontact Holmium:YAG Laser Thermal Keratoplasty United States Phase IIA Clinical Study with a l . . year Follow . . up Douglas D. Koch, MD/ Thomas Kohnen, MD,l Peter J. McDonnell, MD,2 Richard F. Menefee, AAS, 3 Michael J. Berry, PhD3 Purpose: This study was performed to evaluate the safety and effectiveness of noncontact holmium:YAG (Ho:YAG) laser thermal keratoplasty (LTK) for correcting low to moderate hyperopia. Methods: Twenty-eight patients were treated unilaterally to correct low to moderate hyperopia (up to +3.88 diopters [D] refractive error) using simultaneous noncontact delivery of Ho:YAG laser energy. Treatment parameters included one or two symmetric octagonal rings of eight spots per ring with centerline diameters of 6 mm (1 ring) or 6 and 7 mm (2 rings), ten pulses of laser light at 5-Hz pulse repetition frequency, and variable pulse energy, ranging from 208 to 242 mJ. Follow-up was 1 year in 26 (93%) of the 28 patients. Results: At 1 year postoperatively, uncorrected distance visual acuity was improved in all patients. The mean change in subjective manifest refraction (± spherical equivalent [SED was -0.55 ± 0.33 D and -1.64 ± 0.61 D for one- and two-ring treatment groups, respectively, with good stability in the refractive change after approximately 6 months. In the one-ring treatment group (17 eyes), refractive corrections of -0.50 to -1.13 D were achieved in ten eyes (59%), and seven eyes (41%) were unchanged (within ±0.25 D) relative to their preoperative measurements. In the two-ring treatment group, all eight eyes (100%) had substantial refractive corrections (range, -0.75 to -2.50 D). Mean induced refractive astigmatism was 0.25 ± 0.29 D and 0.47 ± 0.53 D for one- and two-ring treatments, respectively. None of the eyes lost two or more lines of spectacle-corrected distance visual acuity. There was no clinically significant change in endothelial cell density with respect to preoperative values. Glare and contrast sensitivity testing indicate that peripheral corneal opacities produced by LTK do not degrade vision. The amount of refractive change in each group was correlated with the amount of laser pulse energy. Conclusions: This initial United States clinical study with 1-year follow-up indicates that noncontact LTK treatment of low hyperopia is safe and effective, providing persistent, though modest, refractive corrections in 59% of the one-ring group and larger, persistent, refractive corrections in 100% of the two-ring group. Ophthalmology 1996; 103: 1525-1536

Originally received: October 31, 1995. Revision accepted: May 30, 1996. I Department of Ophthalmology, Cullen Eye Institute, Baylor College of Medicine, Houston. 2 Department of Ophthalmology, Doheny Eye Institute, University of

Southern California, Los Angeles. 3 Sunrise Technologies, Inc, Fremont, California. Presented in part at the American Academy of Ophthalmology Annual Meeting, Atlanta, OctINov 1995. Supported in part by an unrestricted grant from Research to Prevent Blind-

1525

Ophthalmology

Volume 103, Number 10, October 1996

Laser thermal keratoplasty (LTK) is a promising technique for correcting refractive errors and is currently being evaluated clinically using both contact'-6 and noncontace,8laser energy delivery systems and procedures. In the United States, Food and Drug Administration-monitored studies of contact and noncontact holmium: YAG (Ho: YAG) LTK for correcting hyperopia are in progress. Internationally, a noncontact Ho: YAG LTK clinical study for correcting low hyperopia with a 2-year follow-up has been reported. 8 The purpose of the current article is to report safety and efficacy for the United States noncontact Ho:YAG LTK clinical study based on I-year follow-up measurements. Additionally, the current results are compared with other L TK clinical results for both contact and noncontact modes of laser energy delivery.

Patients and Methods General This clinical study is a prospective, nonrandomized, openlabel investigation of noncontact Ho: YAG LTK for correcting low hyperopia, using preoperative measurements of treated eyes and measurements of fellow, untreated eyes as controls. Fourteen patient treatments were performed in an outpatient setting at each of two sitesCullen Eye Institute, Baylor College of Medicine, Houston, Texas (BCM), and Doheny Eye Institute, University of Southern California, Los Angeles, California (USC). The study had been approved by the institutional review board of each institution before commencement of treatments. Patients Twenty-eight patients (21 women, 7 men; mean age, 54.6 ::t: 7.7 years; range, 38-69 years) were treated by noncontact Ho: YAG LTK unilaterally in their nondominant eyes to correct low to moderate hyperopia (up to +3.88 diopters [D) refractive error). Preoperative entry criteria included patients of at least 21 years of age and preferably older than 45 years; unilateral or bilateral hyperopia (range, 1-5 0 by subjective cycloplegic refraction); best spectacle-corrected distance visual acuity of 20/40 or better in both eyes; refractive and keratometric astigmatism of 1.0 0 or less; stable preoperative refraction; normal corneal topography as assessed with computerized videokeratography; no sightthreatening ophthalmic disease or previous corneal or intraocular surgery; no soft or gas-permeable contact lens ness, Inc, New York, New York, a Deutsche Forschungsgemeinschaft postdoctoral research grant (DFG-Ko 159511-1 and 1-2), Bonn Germany, and by core grant EY03040 from the National Eye Institute, Bethesda, Maryland. Dr. Koch is a paid consultant for Sunrise Technologies. Mr. Menefee and Dr. Berry are employees and stockholders of Sunrise Technologies. Reprint requests to Douglas D. Koch , MD, Cullen Eye Institute, Baylor College of Medicine, 6501 Fannin, NC-200, Houston, TX 77030-3498.

1526

use for at least 2 weeks before baseline measurements and no hard contact lens use for at least 3 weeks with at least two stable refractions and corneal topographic maps taken 2 weeks apart before baseline measurements; and no systemic conditions (e.g., immunocompromised status) or use of systemic medications (such as corticosteroids or antimetabolites) that are likely to affect wound healing. Preoperatively, the mean spherical equivalent of the subjective manifest refraction (SE SMR) was 2.21 ::t: 0.89 o (range, 0.50-3.88 D) for the treated eyes and 1.92 ::t: 0.75 0 (range, 0.88-3.63 D) for the untreated eyes. Laser Thermal Keratoplasty Treatments Noncontact Ho:YAG LTK treatments were performed at the two sites using two matched Sunrise Technologies Corneal Shaping Systems (Fremont, CA). Laser parameters included wavelength (2.13 j.lm), pulse duration (250 j.lseconds), full width at half maximum intensity, pulse repetition frequency (5 Hz), and pulse energy adjustable over the 208- to 242-mJ treatment range used in this study. The device projects a Ho: YAG laser beam treatment pattern of up to eight spots simultaneously on the cornea through a slit-lamp delivery system; a prototype of the delivery system has been described previously.9 Two laser beam treatment patterns were used (Fig 1) in the current study: (1) the one-ring pattern (used to treat 20 eyes with SMR SEs :s 2.5 D) consisted of 8 spots arranged in a symmetric octagonal array of 6-mm centerline diameter, and (2) the two-ring pattern (used to treat 8 eyes with SMR SEs ;::: 2.25 D) consisted of 16 spotsthe same 8 spots at a 6-mm diameter in the one-ring pattern plus 8 additional spots arranged in a symmetric octagonal array of a 7-mm centerline diameter concentric with the one-ring pattern and rotated by 22S. Each spot had a 615-j.lm (USC) or 623-j.lm (BCM) nominal spot diameter (containing approximately 90% of the energy per spot) and a nonuniform energy density distribution within the spot. Individual spot energies were measured with a calibrated joulemeter Model JD 500 (Molectron Detector, Inc, Portland, OR) and were balanced to produce uniform energies within ::t:3 %. Ten laser pulses were delivered sequentially over a 2second period to each treatment ring, yielding a total treatment energy of 2.1 to 2.4 J per ring. Reported pulse energies are average values over the group of ten pulses; individual pulses were uniform within ::t:5% over the pulse train. The treatment pattern (spot centerline) diameters were verified to be 6.0 ::t: 0.1 mm and 7.0 ::t: 0.1 mm at the delivery plane before each patient treatment. The laser pulse energy 1;, was adjusted for each patient treatment to produce a change in central corneal curvature that matched the patient's required refractive correction (from the SE of SMR); values of 1;, were 226 ::t: 9 mJ (range, 208-242 mJ) and 235 ::t: 5 mJ (range, 224-240 mJ) for one- and two-ring treatments, respectively (which corresponds to a range of mean pulse radiant exposures of 8.5 to 10.2 J/cm2). The algorithm relating central corneal shape change to laser pulse energy was determined from 30-day follow-up results at several international clinical sites.

Koch et al . Hyperopia Correction by Noncontact Ho:YAG LTK, US FDA Phase IIA

90

90

\\.\ -,;,,-

180~+4~~++~~~+4~~~ \ \ \ \ \ .... "':/":.,,:/ f J i j

\ e, \" ' '\

..

.:~<",,/

""",\,>-.. _-._- ...

,>:,-

"":>"""e"''',',',,',,',',',',,',,',',',',',',,',',',',e"", ','

.......... """'"

j

......... .

.

o

e,! ,,'

,/

\,e\\ \,,/,,;',/' ",j, ,

\

180~~~~~~~++~~~

\ .... ,.,

-'"

,'

//e/

o

.. \;;/

\""""",\""',,~~~:,'~:.'::

....,',', .... >:

-........ .

-"""

-_ ...... -_ ...... ,.

-"'"

270

270 I I I I I

I

I

012

3

4

Smm

Figure 1. One- and two-ring treatment patterns,

The ideal treatment for this study was defined as centration of the eight laser beam spots with respect to the pupillary center and along the line of sight,1O focusing of each spot on the surface of the cornea, and motionless delivery within each spot during the ten-pulse sequence. Treatments were performed with the physician viewing the cornea with the slit-lamp biomicroscope. Treatment centration was obtained by centering the pattern of the eight red helium-neon (wavelength, 633 nm) laser tracer beams on the iris around the entrance pupil, while the patient monocularly viewed the red light-emitting diode fixation target. Calibrated green helium-neon (wavelength, 543 nm) laser focusing beams were used to focus the laser on the surface of the cornea.

Medications Before treatment, topical anesthetic drops (0.5% proparacaine solution [AlcaineD were administered, beginning at least 20 minutes before treatment, with one drop at 5minute intervals up to a total of four drops. A lid speculum was inserted 5 minutes after the last proparacaine drop was administered, and the eyelids were held open 3 minutes to allow the tear film to dry before treatment. This timing of proparacaine drops and tear film drying was designed to standardize epithelial swelling ll and corneal hydration, variations in which can affect the delivery of Ho:YAG laser light into the corneal stroma. Postoperatively, 0.3% tobramycin antibiotic solution (Tobrex) and diclofenac sodium (Voltaren) were pre-

scribed to be administered to the treated eye four times daily until the epithelium healed. The patients were instructed to take analgesic tablets (plain acetaminophen or acetaminophen with codeine), one or two tablets every 4 hours as needed for pain management. No corticosteroids were used.

Ophthalmic Examinations All ocular examinations were performed by one of two experienced certified ophthalmic technicians and the treating surgeon at each site, using similar examination rooms with 20-foot lanes. Each patient completed the preoperative study questionnaire, and a thorough history and ophthalmic examination were completed. Preoperative ocular measurements included visual acuity (both distance and near, with and without spectacle correction, using standard Snellen charts), refraction (both SMR and subjective cycloplegic refractions obtained using phoropters), keratometry, computerized videokeratography (using EyeSys Corneal Analysis Systems, EyeSys Technologies, Houston, TX), applanation tonometry, ultrasonic pachymetry, contrast sensitivity (using Regan charts with 96%, 50%, 25%, and 11 % contrast levels, Paragon Services, Nova Scotia, Canada), glare disability (using the brightness acuity tester Model 224505 on the medium setting, Mentor-BioRad, Norwall, MA), slit-lamp biomicroscopic evaluation of the cornea, anterior chamber, lens, and fundus, and specular micrography of corneal endothelium (using an Alcon wide-field scanning corneal micro-

1527

Ophthalmology

Volume 103, Number 10, October 1996

Table 1. Postoperative Follow-up Postoperative

No. (%)

1 day 7 days 14 days 2 mos

28 28 26 27 22

3 mos

23 (82)

1 mo

(loa) (loa) (93) (96) (79)

4 mos

20 (71)

6 mos

25 (89)

9 mos

20 (71)

12 mos

26 (93)*

* Loss of follow-up

Results

--------------------------Treatment Precision

Results of examination of slit-lamp biomicroscope photographs, computerized videokeratography eye image, and videotapes of the procedures showed that 20 of the 28 patients received nearly motionless delivery of ten treatment pulses in each ring, producing nearly circular laser treatment spots, whereas the remaining 8 patients received treatment spots that were elongated or otherwise enlarged due to motion. One patient moved during the treatment and received only approximately seven pulses of laser light; this patient initially had some refractive correction in the treated eye, but eventually dropped out of the study.

in two patients.

Refraction scope, Model WFSCM-V, Alcon, at USC, and a KeelerKonan specular microscope, Model SP-I, Konan, Japan, at BCM). The same set of ophthalmic measurements (except for fundus examinations, cycloplegic refractions, and specular micrography, which were performed at reduced frequency), were obtained during follow-up visits on 1, 7, · and 14 days and 1,2,3,4,6,9, and 12 months postoperatively (Table 1). Subjective manifest refraction was not performed on one patient at 1 year. Additionally, at each postoperative visit, slit-lamp examinations included grading of the density of the corneal opacities in the treated spots (0 = undetectable; 1 = barely detectable; 2 = mild opacity; and 3 = dense opacity) and measurement of opacity depth as a percentage of corneal thickness; slitlamp photographs were obtained. On each follow-up visit, patients completed a questionnaire that recorded their subjective evaluations of ocular responses and symptoms (e.g., pain, foreign body sensation, sharpness of vision, sensitivity to bright lights, double vision, tearing, color problem, glare, and diurnal fluctuation of vision) as well as satisfaction with treatment results. Physicians Two physicians performed the 28 procedures-each physician at each site performed 10 procedures with one treatment ring and 4 procedures with two treatment rings.

The mean changes in SE SMR for one- and two-ring treatments were -1.42 :::!::: 0.60 D and -3.97 :::!::: 1.77 D, respectively, at 1 day postoperatively; these regressed to -0.53 :::!::: 0.36 D and -1.92 :::!::: 0.71 D, respectively, at 6 months postoperatively; and then changed minimally over the remaining follow-up visits, resulting in I-year postoperative values of -0.55 :::!::: 0.33 D and -1.64 :::!::: 0.61 D, respectively (Fig 2). The number of eyes with 1 D or less of hyperopia increased from 1 of 18 preoperatively to 6 of 17 at 1 year in the one-ring group and 0 of 8 at both intervals in the two-ring group. The number of eyes with 2 D or less of hyperopia in the two-ring group increased from zero of eight preoperatively to six of eight at 1 year. At 1 year postoperatively, 10 of the 17 eyes treated with one ring (59%) had persistent refractive corrections of

4ar------'I--~--_,-------,------_,__,

I

.

I _.--_._. .._--_... .\ +-_..__-----_. . __...I. ..._ . ._------------- -+----------------_.. _ . ! I I I !

-3.5

f

-3 r--'r--i-----j·--------·~:-

eQ.

-2.5

§. ~

-2

CI)

o

- --

I

I

'

! ;

~ - ·----·--·-··--f- ·- ----~"""-_::.; -··i-'t------

;

f i

I ,

Two ings (n=8)1

- ..- .. -..-. !·-------·-····--I------..·

~ -1.5 ,k-- ............ . . . . .--..... . +..... ..-.. . . . . . . . . . . · · . · . · · 1-···

Data Analysis Visual acuities were averaged using the geometric means. 12 Measurements are typically reported as the mean :::!::: 1 standard deviation (using [n :- 1]112 in ~e denominator of the definition for standard deviation, where n is the number of observations for each measurement) and as the range of all measurements at each follow-up visit. Error bars in figures show the :::!::: 1 standard deviation range. Statistical significances of differences between data sample means are determined by Student's t testsl~; P < 0.05 are considered significant.

1528

CI)

C

-1~-

··+--·-·-..·..·-----..-+---·-·-·--- -- --··[----· ----··t

- ---!:;.==-::::.-=-.~ -! ~=~~==~

-0.5

OL-~~-L~--~~~~~-~~~~

o

90

180

270

360

Day After Treatment Figure 2. Mean changes in spherical equivalents of subjective manifest refraction measurements as a function of mean days after treatment.

Koch et al

Hyperopia Correction by Noncontact Ho:YAG LTK, US FDA Phase IIA

2r--.-------, i -------r------~------_,__, i One Ring n=17) ! !

i

~ .!IQ.

!

·--·----;- -------·--1--- -·-·+-·- ---·---+--

1.5

e.c

Ii

0

~:_+-~.---.-L-----~--~--! Spherical '

CP

c

i

0

Q.

I

E 0

0

.,u~

.

I

'

Keratometry For the patients with the one-ring treatment (n = 18 eyes), the increase in mean keratometric power (calculated from individual means of steepest and flattest meridian powers and corrected to the corneal index of refraction, n = 1.376 14) was 1.66 ± 0.86 D 1day postoperatively, 0.51 ± 0.76 D at 2 months, and 0.39 ± 0.69 D at 1 year. In the two-ring treatment group, mean keratometric power increased 4.92 ± 1.06 D 1 day postoperatively, 2.75 + 1.13 D at 2 months, and 1.51 ± 0.52 D at 1 year.

i

......... ----.- · - --·r-.--.. -.. -· .-.-- ....-l..·.. ·· - --...·-.-.,.-........

0.5

I

-

yylindrical j

i

I

f

I

. -""----'-1-

180

270

····1-,",·--_·_·

~

:

o

90

!

I

.~I

__. ___. ___ ___ _

,

360

Day After Treatment Figure 3. Means of the spherical and cylindrical components of subjective manifest refraction measurements for 17 patients treated with one ring o f spots a s a function of mean days after treatment.

-0.79 ± 0.21 D (range, -0.50 to -1.13 D), whereas 7 eyes (41 %) were unchanged (within ± 0.25 D) from preoperative refractive errors. In the two-ring group at 1 year postoperatively, six of the eight eyes (75%) had substantial stable refractive corrections of -1.92 ± 0.39 D (range, -1.50 to - 2.50 D), whereas two eyes (25%) had low corrections of -0.75 and -0.88 D. Two of the seven unchanged eyes in the one-ring treatment group and both of the eyes with low corrections in the two-ring treatment group were treated with low pulse energies, as described further in the Treatment Algorithms section below. At 1 year postoperatively, control (fellow) eyes had negligible SE SMR changes relative to preoperative values: 0.04 ± 0.50 D (I-ring group) and 0.00 ± 0.35 D (2ring group). Mean refractive cylinder for one- and two-ring treatments (Figs 3 and 4) increased by 0.57 ± 0.64 D and 0.94 ± 1.28 D, respectively, at 1 day postoperatively and regressed to 0.25 ± 0.29 D and 0.47 ± 0.53 D, respectively, at the I-year follow-up visit. At 1 year postoperatively, eight eyes (1 ring, 47%) and five eyes (2 rings, 63%) had increased astigmatism and nine eyes (1 ring, 53%) and 3 eyes (2 rings, 37%) were unchanged (within ± 0.25 D) from preoperative cylindrical errors. The mean change in subjective cycloplegic refraction at the I-year follow-up visit was -0.48 ± 0.38 D (range, 0.25 to -1.13 D; n = 15) for the one-ring group and -1.56 ± 0.58 D (range, -0.75 to -2.25 D; n = 8) for the two-ring group. These values are in close agreement with mean SE SMR corrections, indicating that accommodation was not a factor in assessing refractive outcome for these older patients (mean age, 55 years).

Visual Acuity All 26 treated eyes examined at 1 year postoperatively had increased uncorrected distance visual acuities. Mean uncorrected distance visual acuity (Fig 5) improved from 20/60 preoperatively to 20/30 1 year postoperatively for the one-ring group (n = 18) and from 201125 to 20/50 for the two-ring group (n = 8). The number of eyes with uncorrected distance visual acuity of 20/40 or greater increased from 4 to 11 for the two-ring group (n = 18) and from 0 to 4 for the one-ring group (n = 8). Mean uncorrected near visual acuity (Fig 6) improved from 20/120 to 20/70 in the one-ring group and from 201 250 to 20/120 in the two-ring group. Uncorrected near visual acuities were improved in six and five eyes in the one- and two-ring treatment groups, respectively, unchanged in six and three eyes, respectively, and decreased by one line in six eyes in the one-ring group. There were no changes in mean best-corrected distance and near visual acuities at 1 year postoperatively. At 1 year postoperatively, none of the eyes in either the one-

3.5 3

~ .!IQ.

e. C CP C

1.5

0

Q.

E 0

0

CP > :;:;

1

!

i

I

.

i

i

·- · ··-·······-·- + -

~.-

i

- -- i

0

~

~.5

····t

;

i i

. - ----.. . --.".-, ,-"-

·- 1-

I

1 -·"--···-1- -·---t--

----.---- --.-1 . . ------.----1-.---.-. . -1Spherical -..-... ----1-I __ " ~-----T.

I

..-.

! -- T----·-

0.5

u

£

I Two Ringr (n=8)

-- --+-------.---------.-r-.-.------~ .- -+-- , --

i

2

-

- -_. ._---

2.5

0

t

r--.-------,.--------,--------..--------,--,

i ~yliiidrical -'--

' -"--r -' . . j -····-- 1

i

. ,

--.-- .--------t'~.----~~- --f--~.-

,

o

90

I !

180

; - -·· · .-- +.

..___ ...._.,_.___ ._.~. _ _~. __1 --. - -

270

360

Day After Treatment Figure 4 . Means o fthe spherical and cylindrical components of subj ective manifest refract ion measurements for eight patients treated with two rings of spots as a function of mean days after treatment.

1529

Ophthalmology

i

,

0.8

-------f--···_·_·· ."..,",.." •."."".-,.~.,~" .•-,

Volume 103, Number 10, October 1996

I

t-----r o- oo

~""'."

On~ Ring (n=1~) !,

- - --_ •. . .j"."._,."..........

__oj _____ _

I

I

-_._--_. _-_ .__.. •.,_..,[ ..•....•._•....

i

follow-up visit, one patient lost more than one line, with contrast sensitivity losses of 1.S, 2.S, 2.7, and 2.7 lines at the corresponding contrast levels. Glare Glare test measurements (best-corrected visual acuity using the brightness acuity tester on the medium setting) on 26 patients followed to 1 year postoperatively showed that the mean best-corrected visual acuity was unchanged in the one-ring treatment group and slightly improved in the two-ring treatment group. By 1 week postoperatively, all except one patient in each group had glare scores within one line of preoperative values (2-line decrease in each patient). No patient lost two or more lines of glare acuity at 1 year postoperatively. Slit. lamp Findings

o

90

180

270

360

Day After Treatment Figure 5. Geometric means o f the u ncorrected distance visual acuities (UDV A) for one- and two-ring treatment groups a s a function of mean days after treatment.

or two-ring treatment groups lost two or more Snellen lines of best-corrected distance or near visual acuity. Bestcorrected distance and near visual acuities of all eyes returned to within one line of preoperative values by 2 weeks and 1 week postoperatively, respectively.

Immediately postoperatively, the corneal epithelium was damaged in the treatment spot sites of all patients, showing punctate irregularities with fluorescein staining and pooling. There were no gross epithelial defects. Epithelial healing was completed within 3 to 7 days after treatment. Trace anterior chamber flare was observed in two eyes on the first day after treatment; no significant anterior chamber inflammation occurred in any patient. No changes were noted in the ocular adnexa or in any intraocular structures in the treated eyes. Immediately postoperatively, corneas had depressions in their anterior surfaces at treatment spot sites, accompanied by noticeable central curvature steepenings and, in

Contrast Sensitivity For the one-ring treatment group (n = 18), mean contrast sensitivities decreased (relative to preoperative values) by less than one line at all contrast levels at the I-day followup and then typically increased to equal or better values at later follow-up visits. At 1 year postoperatively, the mean contrast sensitivities were 0.2 ± 1.2, 0.1 ± 1.0, and 0.5 ± 1.3 lines better for 96%, 50%, and 11 % contrast levels, respectively. The postoperative mean contrast sensitivities at the 2S% contrast level remained lower than the preoperative value and finished at -0.2 ± 1.1 lines worse at 1 year postoperatively. At 1 year, the worst contrast sensitivity losses were two lines (1 eye at 96%; and 1 eye at 2S% and 11 % contrast levels); the two patients with these losses did not obtain stable refractive corrections. For the two-ring treatment group (n = 8), mean contrast sensitivities increased (relative to preoperative values) at all follow-up visits for 915%, SO%, and 2S% contrast levels. The mean contrast sensitivity for the 11% contrast level initially decreased by less than one line and then improved by the 30-day follow-up visit, remaining at an increase at later follow-up visits. At 1 year postoperatively, the mean contrast sensitivities were 0.7 ± 1.4, 0.9 ± 2.1,0.7 ± 2.4, and 1.0 ± 2.2 lines betler for 96%, SO%, 2S%, and 11 % contrast levels, respectively. At this

1530

0.4

----,------,-1----,

.----~-----,------,,

I

----I

0.3

i·c

---:

I

:::J

ii E

·uGl

I

0.2

···-······1',-·....

oor'"

~

-1 I

~ z

:)

0.1

o

90

180

270

360

Day After Treatment Figure 6. Geometric means of the uncorrected near visual acuities (UNVA) for one- and two-ring treatment groups as a function of mean days after treatment.

Koch et al . Hyperopia Correction by Noncontact Ho:YAG LTK, US FDA Phase IIA some cases, striations connecting treatment spots. Opacities in each treatment spot were typically 700 to 800 11m in diameter (except when patient and/or physician motion occurred) at the anterior surface and extended in a conical shape to a depth of 40% to 75% of the corneal thickness at each site, depending on treatment energy and motion, if any. Patients typically had grade 2 to 3 (mild to dense) opacities at 1 day. For the one-ring treatment group, mean opacity grades decreased from 2.5 ± 0.5 (between mild and dense opacities) at 1 day to 1.6 ± 0.5 (between barely detectable and mild opacities) at 1 year postoperatively. In the same period, mean opacity depths decreased from approximately 60% to approximately 50% of the corneal thickness. For the two-ring treatment group, mean opacity grades decreased from 2.7 ± 0.4 to 1.6 ± 0.5, and mean opacity depths decreased from approximately 60% to approximately 50% over the same period. Corneal Thickness At all follow-up measurements, the mean central corneal thickness was slightly thinner (range, -2 to -12 11m) compared with the preoperative value. Mean central thickness was 553 ± 31 11m at 1 year postoperatively compared with the preoperative value of 555 ± 31 11m, with no significant differences between one- and two-ring treatment groups.

Intraocular Pressure For one- and two-ring treatment groups, there were small decreases of 1.2 and 1.8 mmHg, respectively, in mean intraocular pressures (lOPs) from preoperative to 1 day postoperative values, followed by fluctuations in mean values up to 3 mmHg lower through the I-year followup period. At 1 year, the mean lOPs for one- and tworing treatment groups were l3.3 ± 3.3 mmHg (range, 820 mmHg) and 15.1 ± 3.6 mmHg (range, 10-20 mmHg), respectively, compared with preoperative values of 15.3 ± 2.3 mmHg (range, 11-20 mmHg) and 14.8 ± 3.4 mmHg (range, 11-20 mmHg) (P > 0.05). Treatment Algorithms This noncontact LTK clinical study on sighted eyes used treatment pulse energy Ep as the primary variable to change the amount of refractive correction, basing initial treatment algorithms for one- and two-ring treatments on 30-day follow-up measurements obtained at international sites. New algorithms were derived in the current study by least-squares fitting of I-year follow-up changes in the SEs of SMRs to linear equations of the form ~ = an + a[Ep, where an and a[ are the intercept and the slope, respectively, of the linear fit. Over the range of one-ring treatment pulse energies (Ep = 208-242 mJ), the one-ring treatment algorithm is: ~SE,

Opacity Depths and Treated Corneal Volumes Slit-lamp biomicroscopic photographs of corneas taken either on the day of treatment (n = 10) or at early followup (1 day [n = 5], 1 week [n = 3], 2 weeks [n = 3], or 1 month [n = 2]) were enlarged and the opacity depths (ODs; units = percentage of paracentral corneal thickness) of the treatment spots were measured. For one-ring treatments (n = 15), the mean OD was 57% ± 11 % (range, 40%-75%); there was no significant difference (P = 0.18; 2-sample equal variance Student's t test) between successful (OD = 60% ± 11 %) and unsuccessful (OD = 52% ± 9%) patients. For two-ring treatments (n = 8), the mean OD was 53% ± 13% (range, 35%-80%); all of these treatments were successful. The treated corneal volume per treatment spot (TCV) was calculated by assuming that each volume has a conical shape with base area A and length (OD/lOO) X paracentral corneal thickness, where the paracentral corneal thickness is taken to be 0.6 mm. This approximation leads to the relation TCV = (Al3) X (OD/lOO) X the paracentral corneal thickness. The mean TCV for the 15 eyes treated with one ring of spots and with early follow-up slit-lamp photographs was 0.048 ± 0.010 mm3 per spot (range, 0.035-0.070 mm 3 per spot); there was no significant (P = 0.35; two-sample equal variance Student's t test) difference between successful (TCV = 0.050 ± 0.011 mm3 per spot) and unsuccessful (TCV = 0.045 ± 0.007 mm3 per spot) patients. The mean TCV for the eight eyes treated with two rings of spots was 0.044 ± 0.009 mm3 per spot (range, 0.035-0.062 mm3 per spot).

SMR2 (diopters) = 3.20 - 0.0l71Ep

Anomalous (5 small and 1 large) refractive corrections were omitted from the fit. The one-ring algorithm for 11 patients has r = 0.48, with a standard error of estimate of 0.19 D. Over the corresponding range of two-ring treatment pulse energies (Ep = 224-240 mJ), the two-ring treatment algorithm is: ~SE,

SMR2 (diopters)

=

14.47 - 0.0685Ep

Anomalous (1 small and 1 large) refractive corrections were omitted from the two-ring fit. The two-ring algorithm for six patients has r2 = 0.75, with a standard error of estimate = 0.25 D. The actual versus calculated refractive correction for both treatment groups is compared in Figure 7. Endothelial Cell Density Specular micrography of the central corneal regions of treated and untreated eyes was performed before treatment and at 3- and 6-month and I-year follow-up visits. Calibrated cell counts were obtained from two or more images to determine the average endothelial cell density (ECD) for each eye and follow-up visit. There was no significant change in ECD for treated or untreated eyes over the follow-up period. For one-ring treatments (n = 15 patients with I-year follow-up measurements), treated eye ECD values were 2810 ± 370 cells/mm2 preoperatively and 2850 ± 300 cells/mm2 at 1 year postoperatively. For two-ring treatments (n = 7 patients with a 1year follow-up), treated eye ECD values were 2860 ±

1531

Volume 103, Number 10, October 1996

Ophthalmology

responded "yes" to the question "Would you have the other eye treated," 2 answered "no," and 3 were uncertain. Four of the five patients who did not respond affirmatively received one-ring treatments and no refractive corrections (within ± 0.25 D).

-2.5

-2

f

J!! c..

e.w 0

Complications -1.5

The only complication was the two-line loss in contrast sensitivity in one two-ring patient.

en

.l!I

a;

C

-1

Discussion

iii ::l U

«

Background

..(1.5

..(1.5

-1

-1.5

-2

-2.5

Calculated Delta SE (Diopters) Figure 7. Scatterplot of actual l-year refractive changes (SE, subjective manifest refraction values) versus calculated changes (from energy-dependent algorithms) for one-ring (circle) and two-ring (square) treatments. The values used in fitting algorithms are within ± 0.5 diopter of the actual = calculated line; the outlier values not used in the fits are still within ± 1.0 diopter of the actual = calculated line.

610 cells/mm2 preoperatively and 2890 ± 340 cells/mm2 at 1 year. The corresponding untreated eye ECD values (n = 22 total patients) were 2830 ± 410 cells/mm2 and 2880 ± 300 cells/mm2 before and 1 year after surgery, respectively. Subjective Patient Responses Postoperative pain was limited to the first day after treatment, with three patients reporting mild-to-moderate intensity. Foreign body sensation was noted in the immediate postoperative period by five patients. Episodic foreign body sensation was experienced by three patients during follow-up examinations in the absence of any epithelial or intraocular abnormalities visible by slit-lamp biomicroscopy. Tearing was reported by four patients for the first 24 hours after treatment, and photophobia was reported by three patients until 2 weeks postoperatively. Patient questionnaires included the question' 'How satisfied are you with the surgery?" Responses were made on a ten-point scale (10 = most satisfied, 1 = very disappointed). Patients reported improved vision in their treated eyes immediately after treatment. The mean response score decreased from 8.4 ± 2.0 (very satisfied) for 1 day postoperatively to 6.7 ± 2.7 (moderately satisfied) for 1 year postoperatively, paralleling the partial loss of refractive correction over time. There were no significant intergroup (1- versus 2-ring treatments) differences in response. At 1 year postoperatively, 21 of 26 total patients completing the I-year follow-up questionnaire

1532

Thermal keratoplasty (TKP) dates back nearly 100 years to Lans' pioneering experiments. 15 Nonlaser modalities that have been investigated and wholly or largely abandoned include the thermokeratophore,16-19 Los Alamos keratoplasty,20 and radial thermokeratoplasty (RTK) as developed by Neumann et a1. 21 The major complications have been regression, poor predictability, and corneal scarring?225 We believe that the primary cause of these complications is excessive corneal heating, producing temperatures much higher than the known human collagen shrinkage temperature of approximately 58° to 60°c.25-27 Laser thermal keratoplasty has been recognized recently as a promising method to produce controlled heating of stromal collagen to modify corneal shape?8 Preclinical studies using a variety of lasers have been reported,29 but to our knowledge only Erbium:glass30 and Ho:YAG (see the following section) lasers have been used thus far in human clinical studies. It is likely that many types of lasers can produce the correct temperature-time history of stromal collagen heating to cause corneal shaping,28,31 but candidate LTK procedures will be successful only if they are free from significant complications and regression of initial refractive changes. Although the detailed failure mechanisms of TKP, RTK, and early CO2 LTK32 are not understood, successful corneal shaping by heat from any source probably requires minimal intervention with a minimum amount of energy. The amount of energy applied must thermally modify a sufficient volume of stromal collagen8 without causing significant long-term biologic, biochemical, and biomechanical responses of the cornea to thermal injury or any other stimulus. Holmium:YAG Laser Thermal Keratoplasty The holmium: YAG laser produces light at a wavelength of approximately 2.1 /-lm. Corneal absorption at this wavelength (assuming that the cornea is 78% water by weight) occurs with a temperature-dependent absorption coefficient at (base e) ranging from 21 to 19 cm- I for the temperature range 35° to 75°C,33 corresponding to an average penetration depth (i.e., the absorption pathlength over which approximately 63% of the light entering the cornea is absorbed) of 480 to 530 /-lm (approximately 80% to 90% of the paracentral corneal thickness), This

Koch et al . Hyperopia Correction by Noncontact Ho:YAG LTK, US FDA Phase lIA range of absorption coefficients, combined with the correct laser pulse waveform and irradiance, is useful to produce intrastromal heating without damaging the corneal endothelium. In contrast to non-LTK and CO 2 LTK, which rely on thermal diffusion to transport heat into the stroma through overheated epithelium, Ho:YAG LTK typically generates more moderate and uniform distributions of temperature deep within the stroma after each laser pulse, thereby minimizing anterior stromal injury. Two types of pulsed Ho:YAG LTK procedures are currently being used to treat hyperopia: (1) a contactmode procedure for sequential delivery of laser energy pulses into individual pre-marked corneal spots using a hand-held fiber optic/focusing optics probe,I-6 and (2) a noncontact-mode procedure for simultaneous delivery of laser energy pulses into a symmetric octagonal pattern of eight laser spots imaged onto the cornea using a slit-lamp delivery systemY The two procedures produce different corneal temperature-time-space distributions: the contact-mode procedure probably heats stromal collagen to a higher average temperature due to the delivery of approximately twice as much energy per spot (19 mJ X 25 pulses versus 24 to 30 mJ X 10 pulses) at three times the pulse repetition frequency (15 versus 5 Hz) in a higher irradiance (strongly versus weakly focused) geometry.

Safety An earlier phase 1 study established the safety of non contact Ho:YAG LTK in a group of ten poorly sighted patients, using three measures (ECD, lOP, and central corneal thickness) to characterize possible adverse effects.7 Additionally, a clinical study on 17 sighted patients treated with one ring of eight spots and followed to 2 years postoperatively demonstrated that no patient lost two or more lines of best-corrected visual acuity and that, at earlier follow-up times, no significant changes in lOP and central corneal thickness occurred. 8 The current study amplifies those results for two groups (1- and 2-ring treatments) followed to 1 year postoperatively: for all safety measures (ECD, lOP, central corneal thickness, and bestcorrected visual acuity), there were no significant postoperative versus preoperative changes in either group, and there were no other adverse effects that raise safety concerns. Patients in the two-ring group were treated with approximately twice the total delivered energy compared with those who were treated previously with noncontact Ho:YAG LTK; the two-ring group results therefore provide a more stringent test of safety issues.

Efficacy The current study generally confirms the efficacy of noncontact Ho:YAG LTK previously observed for one-ring treatments. At 1 year postoperatively, the current onering group (n = 17) obtained a mean refractive correction (~SE, SMR) of -0.55 ::t: 0.33 D compared with the previous one-ring group (n = 16) value of -0.83 ::t: 0.65 D.8 Although the patients currently in the one-ring group were treated with greater mean pulse energy (228 ::t: 6 mJ ver-

sus 179 ::t: 13 mJ for the previous group8), they received substantially less refractive correction. Examination of patient, laser treatment, and procedure variables for the previous and current groups does not show clear differences that may have affected refractive outcome. The seven patients in the current one-ring treatment group who obtained little or no refractive correction were either treated with low energy (n = 2 cases treated with 208 mJ who obtained only -0.25 D correction) or had significantly (P = 0.033 ; two-sample equal variance Student's t test) shallower opacity depths (for 5 patients, the mean OD = 47% ::t: 6%) compared with patients with successful treatments (for 10 patients, the mean OD = 60% ::t: 11%). The efficacy of two-ring treatments in the current study is very encouraging. At 1 year postoperatively, the tworing group (n = 8) had a mean refractive correction (~SE, SMR) of -1.64 ::t: 0.61 D, three times the corresponding value for the one-ring group. In addition, six of the eight patients in this group had substantial corrections (range, -1.5 to -2.5 D); the remaining two patients had modest corrections (-0.75 to -0.88 D) associated with low treatment energies. Even though these two-ring treatments are satisfactory, a bilateral treatment study (in which both patient eyes were treated identically except for the interring rotation) has demonstrated that larger (and better visual quality) two-ring treatments can be obtained using treatment spots located on radials, rather than staggered as shown in Figure 1. 34 Radial treatment patterns will be used in future two-ring treatments for this United States noncontact Ho:YAG clinical study.

Predictability and Stability Long-term (at least 2 years postoperatively) noncontact Ho:YAG LTK data currently are being obtained to construct algorithms that have predictive value. Too few follow-up data are currently available to assess the effects of patient variables, such as age, sex, and corneal structure (thickness, curvature, diameter, etc.), and properties (hydration, etc.) on refractive outcome. At this stage of noncontact Ho:YAG LTK device and procedure development, it is worth noting that short-term (e.g., I-month postoperative) refractive changes regress in part, but that refractive and topographic stabilization35 generally occurs within 6 to 9 months postoperatively to yield some level of improved vision in almost all patients. The long-term stabilities of one-ring treatments by both contact-mode and noncontact-mode Ho: YAG LTK procedures have been discussed elsewhere. 8 The contact-mode mean refractive changes (~SE, SMR values) regressed from approximately -2.8 and -3.5 D at the I-month follow-up for 7.0 mm (3 patients) and 6.5 mm (13 patients) diameter treatments, respectively, to approximately -1.1 D at the 2-year follow-up for both treatments, yielding a regression of approximately 60% to 70% from 1 month to 2 years postoperatively (unpublished data, Thompson VM; presented as a paper at the American Academy of Ophthahpology Annual Meeting, San Francisco, October 1994). In addition, 24% of the contact-

1533

Ophthalmology -7

C~ntact ~.5/9m~ (n=~) i

-6

..

~ CI) Q.

-

-4

UJ

-3

C w ~ CI)

C

I

-r~~

··r-: ..... ' ,

... _L. _____.......____ .L ...__

-5

0

Volume 103, Number 10, October 1996

. . ~j. . .....~.~. ····-··i··~~ . . . . ·f-······ C;:ontaq't 7/9m".. (n=:n i

'

!

i ·=""'t"~::::::t,:::::·--··-·-;;;:~::::--·--·--··t------·--,-

-2 -1

O~~~~-L~~~~~~-L~~~~~~~

o

3

6

9

12

15

18

21

24

Month After Treatment Figure 8. Comparison of mean refraction changes (~standard error, subjective manifest refraction values) for contact laser thermal keratoplasty treatments using 6.5- and 9.0-mm and 7.0- and 9.0-mm diameter rings and for United States noncontact laser thermal keratoplasty treatments using 6- and 7 -mm diameter rings as a function of months after treatment.

mode-treated eyes (both 1- and 2-ring treatments) lost two or more lines of best-corrected visual acuity. For the noncontact~mode procedure, the mean refractive change regressed from -1.14 to -0.79 D (31 % regression) over the same period, and none of the treated eyes lost two or more lines of best~corrected visual acuity.s A subgroup (11 patients who obtained more than -0.25 D of refrac~ tive correction at the 2-year follow~up) of the patients who received the noncontact~mode procedure regressed from -1.35 to -1.07 D (21 % regression) over the same period, ending at approximately the same correction as those who underwent the contact-mode procedure, even though they received much less treatment energy and lower-temperature stromal collagen heating (see above). We do not have the individual patient data needed to do a comparable subgroup analysis of the Summit study. Analogous results have occurred with two~ring contact and noncontact LTK. Figure 8 shows the corresponding refractive power changes (L~SE, SMR values) for two sets of two~ring contact~mode LTK treatments (unpub~ lished data, Thompson VM; presented as a paper at the American Academy of Ophthalmology Annual Meeting, San Francisco, October 1994) (using 2 rings of 8 spots at 6.5- and 9.0-mm [3 patients] and 7.0~ and 9.0~mm [3 patients] centerline diameters) and for 'our two~ring noncontact~mode treatments (using 2 rings of 8 spots at 6~ and 7-mm centerline diameters [8 patients]). The contact~ mode treatments produced large, 1~month postoperative mean refractive changes of -4.8 to -6.5 D, regressing approximately 70% to 75% to 2-year postopefative mean refractive changes of approximately -1.5 to -1.6 D. For

1534

the noncontact~mode procedure, the mean refractive change regressed from -3.20 to -1.64 D (49% regres~ sion) from 1 month to 1 year postoperatively. Extrapola~ tion of noncontact Ho:YAG LTK treatment results indi~ cates that patients who undergo the two-ring treatment may have approximately -1.5 D mean refractive change at 2 years postoperatively (i.e., the same long-term result as for contact~mode Ho:YAG LTK two~ring treatments) . The similarities between contact~ and noncontact~mode long~term results for similar Ho:YAG LTK treatment ge~ ometries are striking . The mechanism(s) responsible for LTK regression cur~ rently are being investigated. From histologic studies on human and rabbit corneas, it is now known that the noncontact Ho:YAG LTK treatment conditions used in the current study generate considerable damage to epithelial cells and keratocytes, but not to endothelial cells?6 It is also known, at least in a rabbit model, that the current treatment conditions: (1) provoke procollagen synthesis by fibroblastic keratocytes, leading to stromal remodeling; and (2) produce anterior corneal surface irregularities, leading to epithelial hyperplasia. The amount of wound~ healing response that occurs in patients under the current treatment conditions is not known. Milder (i.e., lower total-delivered energy per spot) noncontact Ho:YAG LTK treatment conditions currently are being developed in new clinical trials. Although the current treatments yield a tolerable amount of regression, better and more stable treatments involving lower temperature/shorter duration heating of stromal collagen in a variety of treatment pattern geometries have been found. 34

Extension and Refinement of Hyperopia Correction In the current study, we achieved refractive corrections of up to + 2.5 D of hyperopia. Approximately 80% of adults with hyperopia have "low hyperopia," requiring corrections of 3.0 D or less. 3 ? In other noncontact Ho: YAG LTK studies in progress, the range of hyperopia correction with long~term stability is being extended to 3.0 D and more by using three rings of 8 and/or 16 treatment spots at a fixed pulse energy, but variable treatment ring diameters to "titrate" the keratorefractive power change. Other refinements of the treatment algorithms (varying treatment pattern geometries and pulse energies) currently are being developed at several international clinical sites. Procedural refinements also are being evaluated in ongoing clinical studies to improve the reproducibility of motion control, focusing, and other aspects of the treatment. Studies also are under way to assess the efficacy of retreatment of patients who have residual hyperopic refractive errors after initial LTK procedures. The amounts of corneal stromal collagen that have been modified by the current one~ and two~ring procedures are small compared with the amount of peripheral collagen that is available to be treated. We therefore anticipate that it will be possible to increase the refractive change by retreating with additional rings of

Koch et al . Hyperopia Correction by Noncontact Ho:YAG LTK, US FDA Phase IIA spots interspersed between the original ring(s) at 6- and 7mrn diameters or at other diameters. The LTK procedure is simple, fast, safe, and appealing to both patient and physician. If the magnitude of longterm refractive correction and the percentage of successful treatments can be increased, LTK could become the method of choice for correcting low to moderate hyperopia. Acknowledgments. The authors thank the sponsors of this study and David Hennings, MS, for apparatus design and improvements; Elizabeth Haft, COT, JoLene Carranza, COT, Belquiz A. Nassaralla, MD, Murilo V. Domingues, MD, Jenny Garbus, BS, Laurie LaBree, MS, David Klein, PhD, Pamela Buckman, RN, MS, John Carlow, EdD, Donald Guthner, MS, and Carol Kim, BA, for their administrative and clinical assistance in this study; and Giorgio Dorin, BS (PI in Italy), Nick Lurowist, MS, Bruce Sand, MD, and Arthur Vassiliadis, PhD, for their clinical advice and encouragement.

References 1. Durrie DS, Seiler T, King MC, et al. Application of the holmium:YAG laser for refractive surgery. Proc SPIE 1992; 1644:56-60. 2. Seiler T. Ho:YAG laser thermokeratoplasty for hyperopia. Ophthalmol Clin North Am 1992;5:773-80. 3. Thompson VM, Durrie DS, Hunkeler JD, et al. Application of the Ho:YAG laser for refractive surgery: an update of clinical progress. Proc SPIE 1993; 1877:52-6. 4. Thompson VM, Seiler T, Durrie DS, Cavanaugh TB. Holmium:YAG laser thermokeratoplasty for hyperopia and astigmatism: an overview. J Refract Corneal Surg 1993; 9(Suppl):S134-7. 5. Durrie DS, Schumer DJ, Cavanaugh TB. Holmium:YAG laser thermokeratoplasty for hyperopia. J Refract Corneal Surg 1994; 1O(Suppl):S277 -80. 6. Thompson V. Ho: Y AG laser thermokeratoplasty for correction of astigmatism. J Refract Corneal Surg 1994; 10:293. 7. Ariyasu RG, Sand B, Menefee R, et al. Holmium laser thermal keratoplasty of 10 poorly sighted eyes. J Refract Surg 1995; 11:358-65. 8. Koch DD, Abarca A, Villarreal R, et al. Hyperopia correction by noncontact holmium: YAG laser thermal keratoplasty: clinical study with 2-year follow-up. Ophthalmology 1996; 103 :731-40. 9. Parel 1M, Ren Q, Simon G. Noncontact laser photothermal keratoplasty. Biophysical principles and laser beam delivery system. J Refract Corneal Surg 1994; 10:511-8. 10. Uozato H, Guyton DL, Waring GO III. Centering corneal surgical procedures. In: Waring GO III, ed. Refractive Keratotomy for Myopia and Astigmatism. St. Louis: MosbyYear Book, Inc, 1992;491-505. 11. Herse P, Siu A. Short-term effects of proparacaine on human corneal thickness. Acta Ophthalmol 1992;70:740-4. 12. Brown BW Jr, Hollander M. Statistics: A Biomedical Introduction. New York: John Wiley & Sons, 1977; chapts. 4-5. 13. Holladay JT, Prager TC. Mean visual acuity. Am J Ophthalmol 1991; 111:372-4. 14. Mandell RB. Corneal power correction factor for photorefractive keratectomy. J Refract Corneal Surg 1994; 10:125-8. 15. Lans LJ. Experimentelle Untersuchungen tiber Entstehung von Astigmatismus durch nicht-perforirende Corneawunden. Graefes Arch Clin Exp OphthalmolI898;45:117-52. 16. Gasset AR, Kaufman HE. Thermokeratoplasty in the treatment of keratoconus. Am J Ophthalmol 1975;79:226-32.

17. Arentsen 11, Laibson PR. Thermokeratoplasty for keratoconus. Am J Ophthalmol 1976;82:447-9. 18. Aquavella JV, Smith RS, Shaw EL. Alterations in corneal morphology following thermokeratoplasty. Arch Ophthalmol 1976; 94:2082-5. 19. Keates RH, Dingle J. Thermokeratoplasty for keratoconus. Ophthalmic Surg 1975;6:89-92. 20. Rowsey 11, Doss JD. Preliminary report of Los Alamos keratoplasty techniques. Ophthalmology 1981;88:755-60. 21. Neumann AC, Fyodorov S, Sanders DR. Radial thermokeratoplasty for the correction of hyperopia. J Refract Corneal Surg 1990;6:404-12. 22. Fogle JA, Kenyon KR, Stark WJ. Damage to epithelial basement membrane by thermokeratoplasty. Am J Ophthalmol 1977;83:392-401. 23. McDonnell PJ, Garbus J, Romero N, et al. Electrosurgical keratoplasty: clinicopathologic correlation. Arch Ophthalmol 1988; 106:235-8. 24. Feldman ST, Ellis W, Frucht-Pery J, et al. Regression of effect following radial thermokeratoplasty in humans. J Refract Corneal Surg 1989;5:288-91. 25. Feldman ST, Ellis W, Frucht-Pery J, et al. Experimental radial thermokeratoplasty in rabbits. Arch OphthalmoI1990;108:9971000.

26. Stringer H, Parr J. Shrinkage temperature of eye collagen. Nature 1964;204: 1307. 27. Soergel F. Biomechanische Charakterisierung der menschlichen Homhaut mit dynamisch-mechanischer Spektroskopie. Dissertation. Ulm, Germany: University of Ulm, 1994. 28. Mainster MA. Ophthalmic applications of infrared lasersthermal considerations. Invest Ophthalmol Vis Sci 1979; 18:414-20. 29. Koch DD, Berry MJ, Vassiliadis AJ, et al. Noncontact holmium:YAG laser thermal keratoplasty. In: Salz 11, ed. Corneal Laser Surgery. St. Louis: Mosby-Year Book, Inc, 1995;247-54. 30. Kanoda AN, Sorokin AS. Laser correction of hypermetropic refraction. In: Fyodorov SN, ed. Microsurgery of the Eye: Main Aspects. Moscow: MIR Publishers, 1987; 147-54. 31. Sand BS. Method for collagen treatment. United States Patent Number 4, 976,709 (Dec. 11, 1990). 32. Peyman GA, Larson B, Raichand M, Andrews AH. Modification of rabbit corneal cutvature with the use of carbon dioxide laser bums. Ophthalmic Surg 1980; 11:325-9. 33. Jansen ED, van Leeuwen TG, Motamedi M, et al. Temperature dependence of the absorption coefficient of water for midinfrared laser radiation. Lasers Surg Med 1994; 14:258-68. 34. Koch DD, Villarreal R, Kohnen T, et al. Intrastromal procedures: the use of the noncontact holmium:YAG laser for correction of hyperopia-Sunrise Technologies experience. In: Sher NA, ed. Refractive Surgery of Hyperopia and Presbyopia. New York, Tokyo: Igaku-Shoin Medical Publishers (in press). 35. Kohnen T, Husain SE, Koch DD. Corneal topographic changes after noncontact holmium:YAG laser thermal keratoplasty to correct hyperopia. J Cataract Refract Surg 1996; 22:427 - 35. 36. Koch DD, Kohnen T, Anderson J, et al. Histological changes and wound healing response following 10-pulse noncontact holmium: Y AG laser thermal keratoplasty. J Refract Surg 1996; 12:621-34. 3':1. Leibowitz H, Krueger DE, Maunder LR, et al. The Framingham Eye Study Monograph. VIII. Visual acuity. Surv Ophthalmol 1980;24(Suppl):472-9.

1535

Ophthalmology

Volume 103, Number 10, October 1996 Discussion

by Theo Seiler, MD, The correction of hyperopia is one of many problems in refractive surgery that is not yet solved. Several attempts have been made in the past, including keratomileusis, therrnokeratoplasty, hexagonal keratotomy, and automated lamellar keratotomy, none of which has gained wide clinical acceptance. Laser thermokeratoplasty (LTK) is a new approach with the advantage of being less invasive than conventional thermokeratoplasty. On the other hand, as after conventional therrnokeratoplasty, significant regression was anticipated. I•2 The current study seems, at first glance, to demonstrate that this strong regression limits the value of LTK. However, after analyzing the regression curve, it looks exponential with a half time of approximately 80 days. This means that, on average, after three half times (approximately 8 months), a stabilization has occurred. The half time of 80 days may reflect a biomechanical constant of the human cornea because we found an average half time of 78 days when analyzing the regression after LTK for hyperopic astigmatism (unpublished results). However, despite the probably stable refractive effect of +2.5 diopter at 1 year, we should not forget that approximately 35% of the patients have to be considered as nomesponders of hyperopic LTK. The authors should be encouraged to further follow the From the Department of Ophthalmology, TU Dresden, Germany.

1536

PhD

patients to verify the stability of the refractive outcome after 9 months. The good news is that LTK seems to be a safe procedure because none of the patients lost a significant amount of best spectacle-corrected visual acuity. We do not yet know whether hyperopic photorefractive keratectomy may become another alternative in the' 'hyperopia correction jungle"; however, our patients currently prefer LTK compared with hyperopic photorefractive keratectomy because of less halo problems. This is probably due to the larger optical zone (6 mm after LTK versus 3-4 mm after photorefractive keratectomy). In summary, the current article sets a landmark in refractive surgery because the authors demonstrated in a prospective study that holmium: Y AG LTK is a safe procedure used to correct hyperopia up to 2.5 diopters. However, there was a nonresponder rate of 25%. References 1. Neumann AC, Sanders D, Raanan M, DeLucca M. Hyperopic therrnokeratoplasty: clinical evaluation. J Cataract Refract Surg 1991; 17:830-8. 2. Seiler T. Ho:YAG laser therrnokeratoplasty. Ophthalmology Clinics of North America 1992;5:773-80.