Multifocal Corneal Topographic Changes With Excimer Laser Photorefractive Keratectomy Hamilton
Moreira, MD; Jenny J. Garbus; Armand Fasano, MS; Martha Lee, PhD;
Excimer laser photorefractive keratectomy can flatten the central cornea, thereby eliminating myopic refractive errors; in older patients, however, presbyopia limits satisfaction. Computer-assisted topographic analysis of corneas after refractive surgery indicates that a minority of patients achieve a multifocal lens effect, such that they maintain reasonable acuity \s=b\
range of defocus. We have purposeto create a multifocal refractive effect and have analyzed the subover a
fully attempted
sequent topographies quantitatively to determine if multifocality corneas
spheres,
was achieved. In not operated on and plastic hemi-
a fairly small range of corneal powers is observed; the range of powers is increased after a monofocal ablation. After multifocal ablations, a greater spread of surface powers is observed, often with a bimodal distribution, indicative of an apparent multifocal effect. These observations suggest that in some patients
undergoing photorefractive keratectomy myopia, it may be possible to reduce symptoms of presbyopia, although a decrease in image contrast or monocular diplopia may complicate this approach. (Arch Ophthalmol. 1992;110:994-999)
for
of the results of refractive such as radial have docu¬ mented the lack of direct correlation be¬ tween changes in visual acuity, cycloplegic refraction, and keratometric mea¬ surement.1 Using computer-assisted analysis of corneal topography after ra-
Quantitative analysis surgical procedures keratotomy
Terrance N.
flattening.2
The regional variation in corneal power appeared to explain how a multifocal lens effect could be achieved. By contrast, patients with large homogeneous (in terms of corneal power) central corneas did not exhibit this multifocal effect, an observation that has been confirmed by others.3 These inadvertently created multifocal lens effects in corneas after radial keratotomy, however, at least in some pa¬ tients, may produce symptoms of visual distortion or monocular diplopia.2·3 We were intrigued by the possibility that multifocal corneas could be pur¬ posefully created with superficial abla¬ tions using the excimer laser. As cur¬ rently performed for myopia, the sur¬ geon attempts to create relatively large homogeneous zones of central flatten¬ ing, and topographic analysis suggests this is generally achieved.4·5 Theoreti¬ cally, if completely successful, such an approach would result, eventually, in 100% of corrected myopes experiencing presbyopia. This report describes ex¬ perimental creation of multifocal abla¬ tions, designed to correct myopic refrac¬ tive errors while allowing good uncorrected near vision. MATERIALS AND METHODS Ablation Technique
for publication December 16, 1991. From the Doheny Eye Institute and the Department of Ophthalmology, University of Southern California School of Medicine, Los Angeles (Drs Moreira, Lee, and McDonnell and Ms Garbus and Mr Fasano), and VISX Inc, Sunnyvale, Calif (Mr
The laser used is a 193-nm argon-fluoride excimer laser (Twenty-Twenty Excimer La¬ ser, VISX Ine, Sunnyvale, Calif). To correct myopia, this laser uses a computer-controlled iris diaphragm to vary the diameter ofthe ab¬ lation beam so as to ablate more tissue cen¬ trally than peripherally, thereby flattening the cornea. To create monofocal ablations, we performed a single ablation to cause central flattening of 4 diopters (D). To perform a mul¬ tifocal ablation, we used three strategies: two concentric ablations, one of larger diameter (6 mm) and a second additional ablation of smaller diameter (3 mm) (technique 1; Fig 1, top); two ablations of 6 and 3 mm, with the 3-mm ablation decentered inferiorly by 2 mm (technique 2; Fig 2, top); and one single ab¬ lation in which the computer-controlled iris diaphragm is initially fully open to 6 mm and progressively closes until two thirds of the pulses have been delivered, leaving the pe¬ riphery of the ablation zone flattened, but the central area of approximately 3-mm diameter with no intended refractive change (tech¬
nell).
Ablations in Plastic Spheres.—Sixteen plastic hemispheres having known radii of curvature, ranging from steep (6.99 mm) to
See also pp
935, 944, and 977.
dial keratotomy, we observed regional variation in corneal curvature in eyes of patients of the age in which presbyopia would be anticipated, but who had ex¬ cellent uncorrected acuities at distance and near and usually interiorly decentered zones of relatively greater corneal Accepted
Clapham). Reprint requests to Doheny Eye Institute, 1355 San Pablo St, Los Angeles, CA 90033 (Dr McDon-
nique 3; Fig 3, top).
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
Clapham, BSSE;
Peter J.
McDonnell,
MD
medium (7.33 mm) to flat (7.85 mm) (Bronstein Laboratories, Phoenix, Ariz), and cen¬ tral thickness of 3.1 mm, were placed within a special holder to facilitate centration and were examined using the Corneal Modeling System (CMS; Computed Anatomy, New
York, NY),
a
keratoscope-based computer-
assisted topography analysis system. The back surfaces of the hemispheres were painted black to provide contrast for the videokeratoscope ring images. These were di¬ vided into four groups of four hemispheres each and treated as follows: monofocal abla¬ tions of -4.00 D; multifocal ablations with a second 3-mm diameter ablation of -2.00 D centered within the first 6-mm diameter ab¬ lation of -4.00 D (technique 1); multifocal ab¬ lations with a second ablation decentered 2 mm inferiorly (technique 2); and multifocal ablations with a -4.00-D ablation beginning with the diaphragm at 6 mm and interrupted after delivery of two thirds of the pulses
(technique 3).
The corrections in the polymethyl methacrylate hemispheres were measured with a lensometer (AO 012603, American Optical Corp, Buffalo, NY) by an experienced ob¬ server
masked
as
to desired refractive
change. The examiner focused the reticle of the lensometer, then measured each sphere once in sequence, and then again. The reader, who did not know whether a single ablation or multiple ablations had been performed, was asked to read each plastic hemisphere in
the lensometer and determine whether there were one or two end points present and the magnitude of these end points. The reader was instructed not to look at the quantitative lensometer reading until the end point had been chosen. The reader was asked about the distinctness of the end points. The plastic spheres were again examined with the CMS to determine the induced refractive change. Ablations in Polymethyl Methacrylate Blocks.—Ablations in polymethyl methacry¬ late blocks, read with the lensometer, are used routinely to verify refractive effect be¬ fore photorefractive keratectomy for myo¬ pia.5 We performed monofocal and multifocal ablations in four polymethyl methacrylate blocks according to each of the following pa¬ rameters: myopic correction of -4.00 D, 6-mm diameter ablation zone; myopic correc¬ tion of -2.00 D, 3-mm diameter ablation zone; and multifocal ablations using four blocks for each of the three strategies as described above. These ablations were scanned with a Sloan (Santa Barbara, Calif) DekTak model 3030 profilometer, which has a resolution of 2 µ . Ablations in Animal Corneas.—The rab¬ bits used in this study were maintained in an¬ imal care facilities fully accredited by the American Association of Laboratory Animal Science, and all studies were in accordance
Fig 1.—Technique 1Top, Photograph of bifocal ablation in plastic hemisphere. Outer myopic (4-diopter [D]) ablation zone diameter is 6 mm, and inner myopic (additional 2-D) ablation zone diameter is 3 mm. Center, Color-coded keratograph of plastic hemisphere dem¬ onstrates fairly uniform surface curvature before excimer ablation (left), and concentric zones of flattening after both myopic ablations (right) using technique 1. Bottom left, Distribution of surface powers of the central 6 mm of the hemisphere illustrated in Fig 1, center, pre¬ operatively (open squares), after single 4-D ablation (6-mm diameter, diamonds), and after second 2-D ablation (3-mm diameter, closed squares). Bottom right, Profile scan through the center of a flat plas¬ tic block after both myopic ablations (6-mm diameter ablation of -4.00 D; 3-mm diameter ablation of -2.00 D) using technique 1.
V
\ ^_ -C^r*
36
37
Diopters
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
Fig 2.—Technique 2. Top, Photograph of bifocal ablation in plastic hemisphere. Outer myopic (4-diopter [D]) ablation zone diameter is 6 mm, and interiorly decentered (additional 2-D) ablation zone diame¬ ter is 3 mm. Center, Color-coded keratograph of plastic hemisphere after both myopic ablations using technique 2. This demonstrates greater flattening in the interiorly decentered ablated area. Bottom left, Distribution of surface powers of the central 6 mm of the hemisphere illustrated in Fig 2, center, preoperatively (open squares), after single 4-D ablation (6-mm diameter, diamonds), and after second 2-D ab¬ lation (3-mm diameter, closed squares). Bottom right, Profile scan through the center of a flat polymethyl methacrylate block after both myopic ablations (6 mm, -4.00 D; 3 mm, -2.00 D) using tech¬ nique 2.
_
36
38
Diopters
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
Fig 3.—Technique 3. Top, Photograph of bifocal ablation In plastic hemisphere. Annular zone of stepped ablation surrounds central re¬ gion whose curvature is unchanged from preoperatively. Center, Color-coded keratograph of plastic hemisphere demonstrates fairly uniform surface curvature before excimer ablation (left) and after ex¬ cimer ablation (right). Bottom left, Distribution of surface powers of the central region of the hemisphere illustrated in Fig 3, center, preoper¬ atively (open squares) and after single 4-diopter ablation interrupted after delivery of two thirds of the pulses (diamonds). Power distribu¬ tion of central 3 mm is also shown (to mimic pupillary constriction); a shift toward higher (preoperative) corneal powers is demonstrated (closed squares). Bottom right, Profile scan through the center of a flat plastic block after myopic ablation using technique 3.
Y
40
42
44
Diopters
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
with the Association for Research in Vision and Ophthalmology Resolution on the Use of Animals in Research. Adult pigmented rab¬ bits were anesthetized with intramuscular ketamine hydrochloride and xylazine hydro¬ chloride. Preoperatively, keratometry and
computer-assisted topographic analysis were performed. The rabbits were then placed under the excimer laser, topical proparacaine was instilled into the eye, and the epithelium remo ved with a blunt spatula. The eye was then fixated using a vacuum fixation ring, and the area for treatment was centered by using the center of the entrance pupil. Im¬ mediately after surgery and at 3 and 6 weeks postoperatively, computer-assisted topo¬ graphic analysis was performed by the same observer masked as to desired refractive change. A drop of balanced salt solution was placed on the corneas immediately before performing the topographic examination. Monofocal (-4.00 D) and multifocal ablations (three strategies described above) were per¬ formed on two eyes each. Data
Analysis
To examine the distribution of central cor¬ neal powers present preoperatively and after monofocal and multifocal ablations, we exam¬ ined the refractive surface as represented by the CMS.6 This instrument has been demon¬ strated to yield accurate and reproducible re¬ sults when measuring steel balls of known ra¬ dii of curvature' and to have an intraobserver variability on the order of 0.25 D with human corneas.8 The CMS uses proprietary soft¬ ware to calculate corneal power at about 7000 points on the corneal surface; this informa¬ tion, expressed in polar coordinates, is placed within a file and used to generate the colorcoded keratographs. To analyze the topo¬ graphic changes quantitatively before and af¬ ter surgery in this study, we accessed data directly from the processed files. The distri¬ bution of refractive powers at all points within a desired radius of the center was graphed. In this manner, the distribution of powers was examined and characterized as unimodal or multimodal and also character¬ ized as to the amount of homogeneity of pow¬ ers by an observer masked as to the exact type of ablation performed. For polymethyl methacrylate hemi¬ spheres, refractive change was measured with the lensometer by a masked observer and compared with the desired change.
RESULTS Lensometer Readings
Monofocal Ablations.—Measured refractive changes in plastic hemispher¬ ical simulated corneas after monofocal ablations agreed closely with the de¬ sired change (4.22±0.08 and 4.31 ±0.06 D [mean ± SD]). Multifocal Ablations.—Using a lens¬ ometer to measure refractive changes after multifocal ablations in plastic hemispheres was difficult because the end points tended to be less distinct than after monofocal ablations. For tech¬ niques 1 and 2, the readings of the large 6-mm diameter zone of the first ablation
showed a greater degree of apparent re¬ fractive effect after the second smaller ablation was performed (4.70±0.27 and 4.60±0.27 D, for techniques 1 and 2, re¬ spectively). Furthermore, the readings of the second end point, which should have been 6.00 D (4.00-D initial ablation and 2.00-D second ablation) also mea¬ sured greater than intended (6.91 ±0.28 and 6.72 ±0.26 D for techniques 1 and 2,
respectively). For technique 3, the masked examiner thought the 3-mm central area gave a more distinct end point with the lensom¬ eter compared with those of techniques 1
and 2, while the outer zone gave a less dis¬ tinct end point that was comparable with that seen with the other techniques. The outer flattened zone measured 4.25 ± 0.19 D, and the central zone, in which no re¬ fractive change was intended, measured 0.22±0.18 D. These differences in SD are not
statistically significant. Computer-Assisted Topographical Analysis
Reproducible topographic maps of simulated corneas were obtained preop¬ eratively. Following monofocal abla¬ tions, central flattening was demon¬ strated. After bifocal ablations, smaller zones of flattening were seen centrally (technique 1; Fig 1, center) or inferiorly (technique 2; Fig 2, center). The corneas on which the treatment was interrupted so as to minimize flattening of the central 3 mm (technique 3) showed a central area unchanged in curvature from preopera¬ tively, surrounded by an annular region of flattening (Fig 3, center). The computer-generated data of the processed files were analyzed quantita¬ tively and displayed graphically in 0.5-D increments (Figs 1 through 3, bottom left). Preoperatively, relatively little variability in refractive power was
found across the central 6 mm. After the monofocal ablation, the flattening pro¬ duced by the laser was reflected by a shift of the curve toward lower dioptric values, and an increase in width of the distribution of powers was observed. After bifocal ablations, a bimodal curve was observed with a wider distribution of the surface powers. In technique 3, when miosis of 3-mm diameter was simulated by graphing only those powers of the central 3 mm, a shift toward higher dioptric powers was observed (Fig 3, bottom left). This new peak had the same power as one of the peaks of the bimodal curve seen with plotting of all points within the central 6 mm and was equal in power to the pre¬ operative cornea. Thus, with this tech¬ nique, the central corneal power was es¬
sentially unchanged.
Profile of the Ablations.—The pro-
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
Diopters
Fig 4.—Top, Distribution of surface powers of
the central 6
mm
of the rabbit
cornea
preop¬
eratively (diamonds) and after single monofocal ablation (6-mmdiameter, squares). Bot¬
tom, Distribution of surface powers of the central 6 mm of the rabbit cornea preopera¬ tively (diamonds) and after multifocal ablation
using technique 3 (squares). Multifocal abla¬ tion results in greater range of corneal pow¬ ers in a bimodal distribution.
file examination of each ablation agreed with the intended approach (Figs 1 through 3, bottom right). With tech¬ niques 1 and 2, the first larger ablation was equal to the single monofocal abla¬ tion of the same diameter and dioptric power. The second, smaller ablation, centered or inferiorly decentered, matched with the single monofocal abla¬ tion with the same parameters. The mul¬ tifocal ablation in which the treatment was interrupted after two thirds of the pulses (technique 3) showed a flattening in the periphery of the ablation zone, but the inner central area had the same cur¬ vature
as
preoperatively.
Ablations in Adult
Pigmented Rabbits
Computer-assisted topographic anal¬ ysis of the photokeratoscope images
6 weeks after monofocal ablation dem¬ onstrated central corneal flattening; af¬ ter multifocal ablations with technique 3, a flattening of the outer portion of the ablation was demonstrated, while the central area curvature remained about the same as preoperatively. The
computer-generated data of the pro¬ cessed files were analyzed quantita¬ tively and displayed graphically in 1.0-D increments. Preoperatively, rela¬ tively little variability in refractive
power was found across the central 6 mm (Fig 4). After monofocal ablation, the flattening produced by the laser was reflected by a shift of the curve toward lower dioptric values, and an increase in width of the distribution of powers was observed (Fig 4, top). Multifocal abla¬ tions performed with technique 3 pro¬ duced a wider distribution of the diop¬ tric values with a bimodal pattern, with one peak similar to the preoperative curvature (Fig 4, bottom). COMMENT
Multifocal contact and intraocular lenses have been developed and clini¬ cally tested as a response to the loss of accommodation with presbyopia or aphakia. While increasing depth of field, such devices can decrease contrast of the retinal image by at least 50% and can decrease the limit of resolution by one or two lines.9 Thus, there may be clinical settings in which these devices cause
patients
to
experience negative
effects from multifocal lenses or to be dissatisfied with their vision. A multifocal lens effect has been cre¬ ated, inadvertently, in some eyes after refractive surgery, such as radial keratotomy.2·3 We have generally noted that patients who experience this after ra¬ dial keratotomy do not complain and can enjoy good distance and near unaided acuities, but may report monocular diplopia when questioned. Maguire and Bourne,3 however, described a patient in whom a multifocal effect was associ¬ ated with complaints of "disabling vi¬ sual distortion." They suggested that their patient may have had a mild form of keratoconus, so that the topographic findings and distortion may have rep¬ resented the response of an abnormal cornea to surgery. Based on their case report, however, we believe a cautious approach to the deliberate creation of multifocal corneas is prudent. Because of a lack of tear film and depth of field of the computer-assisted
topography instrument, we did not rely solely on this instrument to measure the curvature changes, but also performed lensometer and profilometer readings. Our data demonstrate that it is possible purposefully to create multifocal abla¬ tions in plastic hemispheres that simu¬ late corneas and in living rabbit corneas; when measured with the lens¬
ometer, profilometer, and
a
assisted
computer-
topographic analysis instru¬ ment, two refractive end points and a
bimodal distribution of curvatures are demonstrable. With both monofocal and multifocal ablations, however, there is a wider distribution of refractive powers compared with preoperative status, suggesting that excimer ablations might produce degradation of the reti¬ nal image. Although we attempted to create "bifocal" corneas that would have two discrete refractive powers, our results suggest that a "multifocal" effect was actually achieved, reflected by the increased width of distribution of refractive powers of plastic hemi¬ spheres and corneas. In addition, we observed an increased width of distri¬ bution of refractive powers after single
myopic ("monofocal") ablations, some¬ thing that has not previously, to our knowledge, been commented on. The use in this study of 3- and 6-mm ablation zone diameters was arbitrary. Pupillary diameter does appear to be an important variable in creating multifo¬ cal excimer ablations; we demonstrated
large shifts in the relative distributions of surface corneal powers depending on
whether we examined the central 6 mm or central 3 mm of the cornea. Holladay et al10 showed that pupil size affected refractive error after radial keratotomy, with patients who experienced a myopic change with dilation tending to have the best uncorrected visual acuity. These data suggest that measuring pu¬ pillary diameter preoperatively under appropriate lighting conditions to sim¬ ulate near vision (reading) and distance vision will be important to help deter¬ mine the optimal diameters of the abla¬ tion zones. Proper centration of the ablation with respect to the visual axis and center of
the entrance pupil is extremely impor¬ tant for achieving the multifocal effect. Uozato and Guyton11 have demon¬ strated that refractive surgical proce¬ dures are most appropriately centered over the center of the entrance pupil. If the surgeon were to decenter the exci¬ mer laser photoablation in any direc¬ tion, marked visual distortion might re¬ sult, instead of the beneficial multifocal effect desired. Thus, careful centration will be necessary if we are to avoid ir¬ reversible corneal changes leading to decreased contrast sensitivity and dis¬ tortion. Of the strategies employed in this study to create a bifocal effect, we believe one (technique 3) appears most promising: performing an ablation with the diaphragm initially maximally opened, then interrupting the treat¬ ment so as to to leave the central 3 mm of the ablation zone with a curvature unchanged from preoperative status. In this fashion, we can take advantage of pupillary miosis during accommodation that decreases the percentage of light rays reaching the retina that have tra¬ versed the flattened region of the cor¬ nea. Supporting this conclusion, we have demonstrated a shift to steeper values when we compare the central 3 mm with the central 6 mm of cornea using this strategy. In addition, if the patient is unhappy with the refractive outcome of this multifocal procedure because of monocular diplopia or similar problem, this could be converted to a monofocal ablation by completing the treatment of the central 3 mm of the cornea. We believe, however, that the questions that are as yet unanswered about the optical performance of multi¬ focal intraocular lenses and multifocal corneas argue for a conservative ap¬ proach to human clinical trials of exci¬ mer laser ablation for creation of multi¬ focal corneas. This study was supported by a grant from the Autry Foundation, Los Angeles, Calif, and by an unrestricted grant from Research to Prevent Blindness Ine, New York, NY. Mr Clapham is an employee of VISX Inc. None of the other authors have a financial interest in, or have re¬ ceived consultant fees from, VISX Inc. The manuscript was reviewed by Ann Dawson, medical editor.
References 1. Santos VR, Waring GO III, Lynn MJ, Holladay JT, Sperduto RD, PERK Study Group. Relationship between refractive error and visual acuity in the Prospective Evaluation of Radial Keratotomy (PERK) study. Arch Ophthalmol. 1987;105:86-92. 2. McDonnell PJ, Garbus J, Lopez PF. Topographic analysis and visual acuity after radial keratotomy. Am J Ophthalmol. 1988;106:692-695. 3. Maguire LJ, Bourne WM. A multifocal lens effect as a complication of radial keratotomy. Refract Corneal Surg. 1989;6:394-399. 4. McDonald MB, Frantz JM, Klyce SD, et al. One-year refractive results of central photorefrac-
tive keratectomy for myopia in the nonhuman primate cornea. Arch Ophthalmol. 1990;108:40-47. 5. McDonald MB, FrantzJM, Klyce SD, et al. Central photorefractive keratectomy for myopia: the blind eye study. Arch Ophthalmol. 1990;108:799-808. 6. Gormley DJ, Gersten M, Koplin RS, Lubkin V. Corneal modeling. Cornea. 1988;7:30-35. 7. Hannush SB, Crawford SL, Waring GO III, Gemmill MC, Lynn MJ, Nizam A. Accuracy and precision of keratometry, photokeratoscopy, and
Ophthalmol. 1989;107:1235-1239. 8. Bogan SJ, Waring GO III, Ibrahim 0, Drews C, Curtis L. Classification of normal corneal topog-
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
raphy based on computer-assisted videokeratography. Arch Ophthalmol. 1990;108:945-949. 9. Holladay JT, van Dijk HV, Lang A, et al. Optical performance of multifocal intraocular lenses. J Cataract Refract Surg. 1990;16:413-422. 10. Holladay JT, Lynn MJ, Waring GO III, Gemmill M, Keehn GC, Fielding B. The relationship of visual acuity, refractive error, and pupil size after radial keratotomy. Arch Ophthalmol. 1991;109:70-76. 11. Uozato H, Guyton DL. Centering corneal surgical procedures. Am J Ophthalmol. 1987;103:264\x=req-\ 275.
Moreira, MD; Jenny J. Garbus; Armand Fasano, MS; Martha Lee, PhD;
Excimer laser photorefractive keratectomy can flatten the central cornea, thereby eliminating myopic refractive errors; in older patients, however, presbyopia limits satisfaction. Computer-assisted topographic analysis of corneas after refractive surgery indicates that a minority of patients achieve a multifocal lens effect, such that they maintain reasonable acuity \s=b\
range of defocus. We have purposeto create a multifocal refractive effect and have analyzed the subover a
fully attempted
sequent topographies quantitatively to determine if multifocality corneas
spheres,
was achieved. In not operated on and plastic hemi-
a fairly small range of corneal powers is observed; the range of powers is increased after a monofocal ablation. After multifocal ablations, a greater spread of surface powers is observed, often with a bimodal distribution, indicative of an apparent multifocal effect. These observations suggest that in some patients
undergoing photorefractive keratectomy myopia, it may be possible to reduce symptoms of presbyopia, although a decrease in image contrast or monocular diplopia may complicate this approach. (Arch Ophthalmol. 1992;110:994-999)
for
of the results of refractive such as radial have docu¬ mented the lack of direct correlation be¬ tween changes in visual acuity, cycloplegic refraction, and keratometric mea¬ surement.1 Using computer-assisted analysis of corneal topography after ra-
Quantitative analysis surgical procedures keratotomy
Terrance N.
flattening.2
The regional variation in corneal power appeared to explain how a multifocal lens effect could be achieved. By contrast, patients with large homogeneous (in terms of corneal power) central corneas did not exhibit this multifocal effect, an observation that has been confirmed by others.3 These inadvertently created multifocal lens effects in corneas after radial keratotomy, however, at least in some pa¬ tients, may produce symptoms of visual distortion or monocular diplopia.2·3 We were intrigued by the possibility that multifocal corneas could be pur¬ posefully created with superficial abla¬ tions using the excimer laser. As cur¬ rently performed for myopia, the sur¬ geon attempts to create relatively large homogeneous zones of central flatten¬ ing, and topographic analysis suggests this is generally achieved.4·5 Theoreti¬ cally, if completely successful, such an approach would result, eventually, in 100% of corrected myopes experiencing presbyopia. This report describes ex¬ perimental creation of multifocal abla¬ tions, designed to correct myopic refrac¬ tive errors while allowing good uncorrected near vision. MATERIALS AND METHODS Ablation Technique
for publication December 16, 1991. From the Doheny Eye Institute and the Department of Ophthalmology, University of Southern California School of Medicine, Los Angeles (Drs Moreira, Lee, and McDonnell and Ms Garbus and Mr Fasano), and VISX Inc, Sunnyvale, Calif (Mr
The laser used is a 193-nm argon-fluoride excimer laser (Twenty-Twenty Excimer La¬ ser, VISX Ine, Sunnyvale, Calif). To correct myopia, this laser uses a computer-controlled iris diaphragm to vary the diameter ofthe ab¬ lation beam so as to ablate more tissue cen¬ trally than peripherally, thereby flattening the cornea. To create monofocal ablations, we performed a single ablation to cause central flattening of 4 diopters (D). To perform a mul¬ tifocal ablation, we used three strategies: two concentric ablations, one of larger diameter (6 mm) and a second additional ablation of smaller diameter (3 mm) (technique 1; Fig 1, top); two ablations of 6 and 3 mm, with the 3-mm ablation decentered inferiorly by 2 mm (technique 2; Fig 2, top); and one single ab¬ lation in which the computer-controlled iris diaphragm is initially fully open to 6 mm and progressively closes until two thirds of the pulses have been delivered, leaving the pe¬ riphery of the ablation zone flattened, but the central area of approximately 3-mm diameter with no intended refractive change (tech¬
nell).
Ablations in Plastic Spheres.—Sixteen plastic hemispheres having known radii of curvature, ranging from steep (6.99 mm) to
See also pp
935, 944, and 977.
dial keratotomy, we observed regional variation in corneal curvature in eyes of patients of the age in which presbyopia would be anticipated, but who had ex¬ cellent uncorrected acuities at distance and near and usually interiorly decentered zones of relatively greater corneal Accepted
Clapham). Reprint requests to Doheny Eye Institute, 1355 San Pablo St, Los Angeles, CA 90033 (Dr McDon-
nique 3; Fig 3, top).
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
Clapham, BSSE;
Peter J.
McDonnell,
MD
medium (7.33 mm) to flat (7.85 mm) (Bronstein Laboratories, Phoenix, Ariz), and cen¬ tral thickness of 3.1 mm, were placed within a special holder to facilitate centration and were examined using the Corneal Modeling System (CMS; Computed Anatomy, New
York, NY),
a
keratoscope-based computer-
assisted topography analysis system. The back surfaces of the hemispheres were painted black to provide contrast for the videokeratoscope ring images. These were di¬ vided into four groups of four hemispheres each and treated as follows: monofocal abla¬ tions of -4.00 D; multifocal ablations with a second 3-mm diameter ablation of -2.00 D centered within the first 6-mm diameter ab¬ lation of -4.00 D (technique 1); multifocal ab¬ lations with a second ablation decentered 2 mm inferiorly (technique 2); and multifocal ablations with a -4.00-D ablation beginning with the diaphragm at 6 mm and interrupted after delivery of two thirds of the pulses
(technique 3).
The corrections in the polymethyl methacrylate hemispheres were measured with a lensometer (AO 012603, American Optical Corp, Buffalo, NY) by an experienced ob¬ server
masked
as
to desired refractive
change. The examiner focused the reticle of the lensometer, then measured each sphere once in sequence, and then again. The reader, who did not know whether a single ablation or multiple ablations had been performed, was asked to read each plastic hemisphere in
the lensometer and determine whether there were one or two end points present and the magnitude of these end points. The reader was instructed not to look at the quantitative lensometer reading until the end point had been chosen. The reader was asked about the distinctness of the end points. The plastic spheres were again examined with the CMS to determine the induced refractive change. Ablations in Polymethyl Methacrylate Blocks.—Ablations in polymethyl methacry¬ late blocks, read with the lensometer, are used routinely to verify refractive effect be¬ fore photorefractive keratectomy for myo¬ pia.5 We performed monofocal and multifocal ablations in four polymethyl methacrylate blocks according to each of the following pa¬ rameters: myopic correction of -4.00 D, 6-mm diameter ablation zone; myopic correc¬ tion of -2.00 D, 3-mm diameter ablation zone; and multifocal ablations using four blocks for each of the three strategies as described above. These ablations were scanned with a Sloan (Santa Barbara, Calif) DekTak model 3030 profilometer, which has a resolution of 2 µ . Ablations in Animal Corneas.—The rab¬ bits used in this study were maintained in an¬ imal care facilities fully accredited by the American Association of Laboratory Animal Science, and all studies were in accordance
Fig 1.—Technique 1Top, Photograph of bifocal ablation in plastic hemisphere. Outer myopic (4-diopter [D]) ablation zone diameter is 6 mm, and inner myopic (additional 2-D) ablation zone diameter is 3 mm. Center, Color-coded keratograph of plastic hemisphere dem¬ onstrates fairly uniform surface curvature before excimer ablation (left), and concentric zones of flattening after both myopic ablations (right) using technique 1. Bottom left, Distribution of surface powers of the central 6 mm of the hemisphere illustrated in Fig 1, center, pre¬ operatively (open squares), after single 4-D ablation (6-mm diameter, diamonds), and after second 2-D ablation (3-mm diameter, closed squares). Bottom right, Profile scan through the center of a flat plas¬ tic block after both myopic ablations (6-mm diameter ablation of -4.00 D; 3-mm diameter ablation of -2.00 D) using technique 1.
V
\ ^_ -C^r*
36
37
Diopters
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
Fig 2.—Technique 2. Top, Photograph of bifocal ablation in plastic hemisphere. Outer myopic (4-diopter [D]) ablation zone diameter is 6 mm, and interiorly decentered (additional 2-D) ablation zone diame¬ ter is 3 mm. Center, Color-coded keratograph of plastic hemisphere after both myopic ablations using technique 2. This demonstrates greater flattening in the interiorly decentered ablated area. Bottom left, Distribution of surface powers of the central 6 mm of the hemisphere illustrated in Fig 2, center, preoperatively (open squares), after single 4-D ablation (6-mm diameter, diamonds), and after second 2-D ab¬ lation (3-mm diameter, closed squares). Bottom right, Profile scan through the center of a flat polymethyl methacrylate block after both myopic ablations (6 mm, -4.00 D; 3 mm, -2.00 D) using tech¬ nique 2.
_
36
38
Diopters
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
Fig 3.—Technique 3. Top, Photograph of bifocal ablation In plastic hemisphere. Annular zone of stepped ablation surrounds central re¬ gion whose curvature is unchanged from preoperatively. Center, Color-coded keratograph of plastic hemisphere demonstrates fairly uniform surface curvature before excimer ablation (left) and after ex¬ cimer ablation (right). Bottom left, Distribution of surface powers of the central region of the hemisphere illustrated in Fig 3, center, preoper¬ atively (open squares) and after single 4-diopter ablation interrupted after delivery of two thirds of the pulses (diamonds). Power distribu¬ tion of central 3 mm is also shown (to mimic pupillary constriction); a shift toward higher (preoperative) corneal powers is demonstrated (closed squares). Bottom right, Profile scan through the center of a flat plastic block after myopic ablation using technique 3.
Y
40
42
44
Diopters
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
with the Association for Research in Vision and Ophthalmology Resolution on the Use of Animals in Research. Adult pigmented rab¬ bits were anesthetized with intramuscular ketamine hydrochloride and xylazine hydro¬ chloride. Preoperatively, keratometry and
computer-assisted topographic analysis were performed. The rabbits were then placed under the excimer laser, topical proparacaine was instilled into the eye, and the epithelium remo ved with a blunt spatula. The eye was then fixated using a vacuum fixation ring, and the area for treatment was centered by using the center of the entrance pupil. Im¬ mediately after surgery and at 3 and 6 weeks postoperatively, computer-assisted topo¬ graphic analysis was performed by the same observer masked as to desired refractive change. A drop of balanced salt solution was placed on the corneas immediately before performing the topographic examination. Monofocal (-4.00 D) and multifocal ablations (three strategies described above) were per¬ formed on two eyes each. Data
Analysis
To examine the distribution of central cor¬ neal powers present preoperatively and after monofocal and multifocal ablations, we exam¬ ined the refractive surface as represented by the CMS.6 This instrument has been demon¬ strated to yield accurate and reproducible re¬ sults when measuring steel balls of known ra¬ dii of curvature' and to have an intraobserver variability on the order of 0.25 D with human corneas.8 The CMS uses proprietary soft¬ ware to calculate corneal power at about 7000 points on the corneal surface; this informa¬ tion, expressed in polar coordinates, is placed within a file and used to generate the colorcoded keratographs. To analyze the topo¬ graphic changes quantitatively before and af¬ ter surgery in this study, we accessed data directly from the processed files. The distri¬ bution of refractive powers at all points within a desired radius of the center was graphed. In this manner, the distribution of powers was examined and characterized as unimodal or multimodal and also character¬ ized as to the amount of homogeneity of pow¬ ers by an observer masked as to the exact type of ablation performed. For polymethyl methacrylate hemi¬ spheres, refractive change was measured with the lensometer by a masked observer and compared with the desired change.
RESULTS Lensometer Readings
Monofocal Ablations.—Measured refractive changes in plastic hemispher¬ ical simulated corneas after monofocal ablations agreed closely with the de¬ sired change (4.22±0.08 and 4.31 ±0.06 D [mean ± SD]). Multifocal Ablations.—Using a lens¬ ometer to measure refractive changes after multifocal ablations in plastic hemispheres was difficult because the end points tended to be less distinct than after monofocal ablations. For tech¬ niques 1 and 2, the readings of the large 6-mm diameter zone of the first ablation
showed a greater degree of apparent re¬ fractive effect after the second smaller ablation was performed (4.70±0.27 and 4.60±0.27 D, for techniques 1 and 2, re¬ spectively). Furthermore, the readings of the second end point, which should have been 6.00 D (4.00-D initial ablation and 2.00-D second ablation) also mea¬ sured greater than intended (6.91 ±0.28 and 6.72 ±0.26 D for techniques 1 and 2,
respectively). For technique 3, the masked examiner thought the 3-mm central area gave a more distinct end point with the lensom¬ eter compared with those of techniques 1
and 2, while the outer zone gave a less dis¬ tinct end point that was comparable with that seen with the other techniques. The outer flattened zone measured 4.25 ± 0.19 D, and the central zone, in which no re¬ fractive change was intended, measured 0.22±0.18 D. These differences in SD are not
statistically significant. Computer-Assisted Topographical Analysis
Reproducible topographic maps of simulated corneas were obtained preop¬ eratively. Following monofocal abla¬ tions, central flattening was demon¬ strated. After bifocal ablations, smaller zones of flattening were seen centrally (technique 1; Fig 1, center) or inferiorly (technique 2; Fig 2, center). The corneas on which the treatment was interrupted so as to minimize flattening of the central 3 mm (technique 3) showed a central area unchanged in curvature from preopera¬ tively, surrounded by an annular region of flattening (Fig 3, center). The computer-generated data of the processed files were analyzed quantita¬ tively and displayed graphically in 0.5-D increments (Figs 1 through 3, bottom left). Preoperatively, relatively little variability in refractive power was
found across the central 6 mm. After the monofocal ablation, the flattening pro¬ duced by the laser was reflected by a shift of the curve toward lower dioptric values, and an increase in width of the distribution of powers was observed. After bifocal ablations, a bimodal curve was observed with a wider distribution of the surface powers. In technique 3, when miosis of 3-mm diameter was simulated by graphing only those powers of the central 3 mm, a shift toward higher dioptric powers was observed (Fig 3, bottom left). This new peak had the same power as one of the peaks of the bimodal curve seen with plotting of all points within the central 6 mm and was equal in power to the pre¬ operative cornea. Thus, with this tech¬ nique, the central corneal power was es¬
sentially unchanged.
Profile of the Ablations.—The pro-
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
Diopters
Fig 4.—Top, Distribution of surface powers of
the central 6
mm
of the rabbit
cornea
preop¬
eratively (diamonds) and after single monofocal ablation (6-mmdiameter, squares). Bot¬
tom, Distribution of surface powers of the central 6 mm of the rabbit cornea preopera¬ tively (diamonds) and after multifocal ablation
using technique 3 (squares). Multifocal abla¬ tion results in greater range of corneal pow¬ ers in a bimodal distribution.
file examination of each ablation agreed with the intended approach (Figs 1 through 3, bottom right). With tech¬ niques 1 and 2, the first larger ablation was equal to the single monofocal abla¬ tion of the same diameter and dioptric power. The second, smaller ablation, centered or inferiorly decentered, matched with the single monofocal abla¬ tion with the same parameters. The mul¬ tifocal ablation in which the treatment was interrupted after two thirds of the pulses (technique 3) showed a flattening in the periphery of the ablation zone, but the inner central area had the same cur¬ vature
as
preoperatively.
Ablations in Adult
Pigmented Rabbits
Computer-assisted topographic anal¬ ysis of the photokeratoscope images
6 weeks after monofocal ablation dem¬ onstrated central corneal flattening; af¬ ter multifocal ablations with technique 3, a flattening of the outer portion of the ablation was demonstrated, while the central area curvature remained about the same as preoperatively. The
computer-generated data of the pro¬ cessed files were analyzed quantita¬ tively and displayed graphically in 1.0-D increments. Preoperatively, rela¬ tively little variability in refractive
power was found across the central 6 mm (Fig 4). After monofocal ablation, the flattening produced by the laser was reflected by a shift of the curve toward lower dioptric values, and an increase in width of the distribution of powers was observed (Fig 4, top). Multifocal abla¬ tions performed with technique 3 pro¬ duced a wider distribution of the diop¬ tric values with a bimodal pattern, with one peak similar to the preoperative curvature (Fig 4, bottom). COMMENT
Multifocal contact and intraocular lenses have been developed and clini¬ cally tested as a response to the loss of accommodation with presbyopia or aphakia. While increasing depth of field, such devices can decrease contrast of the retinal image by at least 50% and can decrease the limit of resolution by one or two lines.9 Thus, there may be clinical settings in which these devices cause
patients
to
experience negative
effects from multifocal lenses or to be dissatisfied with their vision. A multifocal lens effect has been cre¬ ated, inadvertently, in some eyes after refractive surgery, such as radial keratotomy.2·3 We have generally noted that patients who experience this after ra¬ dial keratotomy do not complain and can enjoy good distance and near unaided acuities, but may report monocular diplopia when questioned. Maguire and Bourne,3 however, described a patient in whom a multifocal effect was associ¬ ated with complaints of "disabling vi¬ sual distortion." They suggested that their patient may have had a mild form of keratoconus, so that the topographic findings and distortion may have rep¬ resented the response of an abnormal cornea to surgery. Based on their case report, however, we believe a cautious approach to the deliberate creation of multifocal corneas is prudent. Because of a lack of tear film and depth of field of the computer-assisted
topography instrument, we did not rely solely on this instrument to measure the curvature changes, but also performed lensometer and profilometer readings. Our data demonstrate that it is possible purposefully to create multifocal abla¬ tions in plastic hemispheres that simu¬ late corneas and in living rabbit corneas; when measured with the lens¬
ometer, profilometer, and
a
assisted
computer-
topographic analysis instru¬ ment, two refractive end points and a
bimodal distribution of curvatures are demonstrable. With both monofocal and multifocal ablations, however, there is a wider distribution of refractive powers compared with preoperative status, suggesting that excimer ablations might produce degradation of the reti¬ nal image. Although we attempted to create "bifocal" corneas that would have two discrete refractive powers, our results suggest that a "multifocal" effect was actually achieved, reflected by the increased width of distribution of refractive powers of plastic hemi¬ spheres and corneas. In addition, we observed an increased width of distri¬ bution of refractive powers after single
myopic ("monofocal") ablations, some¬ thing that has not previously, to our knowledge, been commented on. The use in this study of 3- and 6-mm ablation zone diameters was arbitrary. Pupillary diameter does appear to be an important variable in creating multifo¬ cal excimer ablations; we demonstrated
large shifts in the relative distributions of surface corneal powers depending on
whether we examined the central 6 mm or central 3 mm of the cornea. Holladay et al10 showed that pupil size affected refractive error after radial keratotomy, with patients who experienced a myopic change with dilation tending to have the best uncorrected visual acuity. These data suggest that measuring pu¬ pillary diameter preoperatively under appropriate lighting conditions to sim¬ ulate near vision (reading) and distance vision will be important to help deter¬ mine the optimal diameters of the abla¬ tion zones. Proper centration of the ablation with respect to the visual axis and center of
the entrance pupil is extremely impor¬ tant for achieving the multifocal effect. Uozato and Guyton11 have demon¬ strated that refractive surgical proce¬ dures are most appropriately centered over the center of the entrance pupil. If the surgeon were to decenter the exci¬ mer laser photoablation in any direc¬ tion, marked visual distortion might re¬ sult, instead of the beneficial multifocal effect desired. Thus, careful centration will be necessary if we are to avoid ir¬ reversible corneal changes leading to decreased contrast sensitivity and dis¬ tortion. Of the strategies employed in this study to create a bifocal effect, we believe one (technique 3) appears most promising: performing an ablation with the diaphragm initially maximally opened, then interrupting the treat¬ ment so as to to leave the central 3 mm of the ablation zone with a curvature unchanged from preoperative status. In this fashion, we can take advantage of pupillary miosis during accommodation that decreases the percentage of light rays reaching the retina that have tra¬ versed the flattened region of the cor¬ nea. Supporting this conclusion, we have demonstrated a shift to steeper values when we compare the central 3 mm with the central 6 mm of cornea using this strategy. In addition, if the patient is unhappy with the refractive outcome of this multifocal procedure because of monocular diplopia or similar problem, this could be converted to a monofocal ablation by completing the treatment of the central 3 mm of the cornea. We believe, however, that the questions that are as yet unanswered about the optical performance of multi¬ focal intraocular lenses and multifocal corneas argue for a conservative ap¬ proach to human clinical trials of exci¬ mer laser ablation for creation of multi¬ focal corneas. This study was supported by a grant from the Autry Foundation, Los Angeles, Calif, and by an unrestricted grant from Research to Prevent Blindness Ine, New York, NY. Mr Clapham is an employee of VISX Inc. None of the other authors have a financial interest in, or have re¬ ceived consultant fees from, VISX Inc. The manuscript was reviewed by Ann Dawson, medical editor.
References 1. Santos VR, Waring GO III, Lynn MJ, Holladay JT, Sperduto RD, PERK Study Group. Relationship between refractive error and visual acuity in the Prospective Evaluation of Radial Keratotomy (PERK) study. Arch Ophthalmol. 1987;105:86-92. 2. McDonnell PJ, Garbus J, Lopez PF. Topographic analysis and visual acuity after radial keratotomy. Am J Ophthalmol. 1988;106:692-695. 3. Maguire LJ, Bourne WM. A multifocal lens effect as a complication of radial keratotomy. Refract Corneal Surg. 1989;6:394-399. 4. McDonald MB, Frantz JM, Klyce SD, et al. One-year refractive results of central photorefrac-
tive keratectomy for myopia in the nonhuman primate cornea. Arch Ophthalmol. 1990;108:40-47. 5. McDonald MB, FrantzJM, Klyce SD, et al. Central photorefractive keratectomy for myopia: the blind eye study. Arch Ophthalmol. 1990;108:799-808. 6. Gormley DJ, Gersten M, Koplin RS, Lubkin V. Corneal modeling. Cornea. 1988;7:30-35. 7. Hannush SB, Crawford SL, Waring GO III, Gemmill MC, Lynn MJ, Nizam A. Accuracy and precision of keratometry, photokeratoscopy, and
Ophthalmol. 1989;107:1235-1239. 8. Bogan SJ, Waring GO III, Ibrahim 0, Drews C, Curtis L. Classification of normal corneal topog-
Downloaded From: http://archopht.jamanetwork.com/ by a UQ Library User on 06/17/2015
raphy based on computer-assisted videokeratography. Arch Ophthalmol. 1990;108:945-949. 9. Holladay JT, van Dijk HV, Lang A, et al. Optical performance of multifocal intraocular lenses. J Cataract Refract Surg. 1990;16:413-422. 10. Holladay JT, Lynn MJ, Waring GO III, Gemmill M, Keehn GC, Fielding B. The relationship of visual acuity, refractive error, and pupil size after radial keratotomy. Arch Ophthalmol. 1991;109:70-76. 11. Uozato H, Guyton DL. Centering corneal surgical procedures. Am J Ophthalmol. 1987;103:264\x=req-\ 275.
