Since the first human lamellar refractive keratoplasty procedures were reported by Jose Barraquer in 1964,1'2 surgeons have modified his concepts - especially to reduce moderate to high myopia. Barraquer's technique involved freezing and lathing of the partial thickness corneal disc removed from the patient's eye with a microkeratome. The freezing can damage epithelium, keratocytes, and other constituents of the corneal disc,3 which can prolong healing time.4 Freezing also introduces structural alterations in the tissue and cryolathe,5 thus increasing error in the lathing process. Inaccurate results have plagued cryolathe myopic keratomileusis, with 26% to 87% of eyes6"12 having final refractions within ± 2.00 D of desired result.
To circumvent some of these difficulties, Krumeich and Swinger5 introduced the nonfreeze, planar, lamellar refractive keratoplasty instruments and technique in 1985. Although it avoids cryotrauma to the corneal disc, this technique is complex and surgically demanding, and the accuracy of refractive outcome is not optimal, with the averopic keratomileusis.13 To further simplify keratomileusis, Ruiz proposed the in situ myopic technique (oral communication, American Academy of Ophthalmology, Dallas, Tex, 1986), in which the corneal disc is removed with the microkeratome and a second pass of the microkeratome is made to remove a piano disc from the stromal bed, which creates a refractive flattening of the cornea when the original disc is replaced.
This technique avoids cryotrauma and the need for ancillary instruments, but also lacks an accurate refractive outcome. A series of cases by Bas and Nano14 showed a refractive outcome mthin ± 2.00 D in 63% of eyes. The main problems with all three techniques are surgical complexity, inaccurate instrumentation, and unpredictable outcome (Table 1).
Recognizing the limitations of these techniques, we investigated a new technique of myopic keratomileusis removing stromal tissue from the disc with the argon fluoride (193 nm) excimer laser in human cadaver eyes. We compared the changes in corneal curvature induced by the excimer laser with the BKS, planar, nonfreeze method.
MATERIALS AND METHODS
Sixteen human cadaver eyes had been frozen between 1 and 19 weeks. The eyes were sealed in plastic bags and thawed in warm water. A 10.5% dextran-70 solution was prepared by combining equal parts of Swinger-Kornmehl solution and 6% dextran-70, which was injected both into the vitreous cavity through a 27-gauge needle through the optic nerve and into the anterior chamber through an oblique, self-sealing paracentesis tract initiated 2 to 3 mm behind the limbus. The eyes were submerged in Swinger-Kornmehl solution until the corneas reached a central thickness between 525 and 600 µm (average, 576 ± 24 µm), as measured with the Pachpen ultrasonic pachometer (BioRad, Irvine, Calif). This dehydration generally required 2 to 4 hours.
All eyes were free of corneal disease when inspected by slit-lamp microscopy and were assigned randomly to undergo either the BKS or the excimer laser method of myopic keratomileusis. The procedures were alternated to negate learning bias. Each eye underwent the following steps. The epithelium was removed with a blunt angled scalpel blade. A radial reference mark was placed with methylene blue at the 12 o'clock position of the cornea. The intraocular pressure (IOP) was measured with a pneumotonometer (BioRad, Cambridge, Mass) and 10.5% dextran 70 solution was injected through the optic nerve to attain pressure between 15 and 25 mm Hg. The eye was positioned anatomically in an artificial orbit (Jean-Marie Parel, Miami, Fla), centered under a vertically mounted keratometer (Bausch & Lomb, Rochester, NY), and the preoperative keratometric measurements were recorded.
Summary of Published Results of Myopic Keratomileusis
The initial microkeratome section was performed identically for both groups, utilizing the microkeratome from the BKS 1000® refractive set (Eye Tech, Liechtenstein). The microkeratome was disassembled and cleaned before each procedure. A 4. 14millimeter blade was used for up to six passes before being replaced. (No correlation was noted between the thickness of the resected disc and the number of passes made with the blade.) The 0.30-millimeter depth plate was used in the microkeratome, and the suction ring specified by the BKS-Krumeich nomogram for the preoperative corneal curvature was selected. After assuring proper functioning of the apparatus, the suction ring was centered on the cornea, 800 millibars of suction were applied, IOP was confirmed as 65 mm Hg or greater with the surgical applanation tonometer, and the 9.00millimeter diameter to be resected was confirmed with the applanation lens. The surgeon passed the microkeratome over the cornea at a rate of about 1 mm/sec. The corneal disc diameter was measured with calipers, and the thickness was measured with a thickness gauge (Mititoyo, Tokyo, Japan), The excised corneal disc was then placed in an airtight corneal preservation chamber (Stein way Instrument, Inc, San Diego, Calif).
For the BKS method, the refractive bench apparatus was prepared to achieve a 9.00-diopter myopic correction. The spacing ring was selected based on the BKS-Krumeich nomogram and the thickness of the corneal disc. The #69 die was placed over the spacing ring and the dovetail ring positioned and zeroed. The microkeratome without a depth plate was positioned on the dovetail guide. The guide was turned down until it almost touched the die. A dot of methylene blue was placed on the center of the die, and the dovetail ring with the microkeratome was lowered slowly while the microkeratome was moved back and forth until the blade just grazed the colored dot. The indicator on the refractive bench apparatus was then marked at this position. The dovetail ring was rotated counterclockwise to the appropriate position as determined by the BKS-Krumeich nomogram based on the thickness of the corneal disc. The corneal disc was removed from the preservation chamber and placed epithelial side down on the die. Suction was applied transiently, the retaining ring secured, and the edge of the tissue checked for proper fixation. The refractive stromal excision was done with the microkeratome, with the suction applied. For the laser method, the corneal disc was placed epithelial side down in the corneal preservation chamber and centered with the aid of calipers, under the argon fluoride (193 nm) excimer laser (Excimed® UV200, Summit Technologies, Inc, Boston, Mass). Two HeNe aiming beams were focused on the surface. A 9.00-diopter correction was entered into the computer program and the corneal disc was ablated using the following laser parameters: radiant energy 180 mJ/cm2, repetition rate 10 Hz, diameter of ablation 5.0 mm, estimated ablation rate 0.25 µ/pulse, and a total of 349 pulses.
For both procedures, the corneal disc was replaced on the stromal bed in its original orientation without sutures; the IOP was adjusted by injection of 10.5% dextran-70, if necessary, to achieve an IOP between 15 and 25 mm Hg, and postoperative keratometry was performed. The surface (Bowman's membrane) was moistened with 10.5% dextran 70 and the disc was firmly positioned to achieve uniform tissue contact and regular and reproducible keratometer mires. Keratometric measurements were recorded, and the amount of keratometric flattening induced by the procedure was calculated.
The corneal discs were measured just prior to the refractive cut for evidence of thickness change. The time the discs were stored in the preservation chamber ranged between 3 to 30 minutes. All keratometric measurements were taken by two independent observers, as an average of three observations each. The keratometer mires appeared qualitatively the same for both groups of eyes.
Changes in Corneal Power (D) After BKS* and Excimer Laser† Myopic Keratomileusis in Human Donor Eyes
A total of 16 eyes were used in this study. The excimer laser group and the BKS groups each contained 6 eyes; 2 eyes were used for practice, 1 eye was excluded due to abnormally steep corneal curvature, and 1 eye was excluded due to imprecise zeroing of the BKS refractive bench apparatus.
The first three corneas that underwent laser ablation developed gas pockets in the stroma during the ablation, and in one of the corneal discs a linear rupture of the ablated surface developed over a gas pocket. None of the other corneas developed gas pockets. When a fresh corneal disc was lasered, this phenomenon did not occur.
The data acquired from each procedure are summarized in Table 2. The preoperative average keratometry was 43.50 ± 1.20 D for the BKS group and 45.00 ± 1.50 D for the laser group. The diameter of the corneal disc averaged 9.00 mm (SD, 0.13). The average thickness of the corneal disc was less in the BKS group (238 µ) than in the laser group (267 µ); we did not think this difference should affect the refractive cut and amount of corneal flattening. The corneal thickness changes while the corneal disc was stored in the corneal preservation chamber averaged -4% (range, 0% to -8%). The amount of thinning was not proportional to the time the disc was stored in the preservation chamber.
The average corneal flattening with the BKS method was 3.80 ± 1.30 D and with the laser method 5.80 ± 1.00 D, a statistically significant difference (p < .05, Student's t-test). The average preoperative corneal astigmatism for all eyes was 0.84 ± 0.73 D; postoperatively it was 1.33 ± 1.13 D. Induced astigmatism, as determined by vectoral analysis, was relatively low (BKS group 0.40 ± 1.70 D, laser group 0.60 ± 1.30 D) and was not significantly different between the two procedures (p < .3, Student's t-test). There was no consistent axis of resultant astigmatism in either group.
Neither technique achieved the desired 9.00 D of flattening. We found a mean difference of 2.00 D in the amount of central corneal flattening obtained between mechanical and laser tissue removal (Table 2). The reasons for this difference are unknown. The algorithm used in the BKS method may aim for less keratometric flattening than that used in the laser method. Also, using the microkeratome to remove stroma involves mechanical forces that could distort and drag the tissue, altering results, in addition, error could be introduced by the tolerances inherent in mechanical devices.
In this study, the resection or ablation zone diameters of the two techniques were not the same; BKS - slightly less than 5.5 mm, and excimer laser - 5.0 mm. In addition, the mean thickness of the BKS corneal discs was less than the laser corneal discs, which further reduced the relative diameter of the resected tissue and could produce a smaller than expected change in the keratometric flattening in the BKS group. Refinements in the algorithm used with the laser could probably be made.
There are limitations inherent in corneal surgery on a cadaver eye.15 Freezing, thawing, and dehydration of the corneas can induce structural changes in the corneal tissue and influence its response to surgical manipulation. The cutting characteristics of the microkeratome and laser ablation rate may have been altered by these factors. Variations in corneal thicknesses could also affect the results. We attempted to minimize the error involved in this type of comparison study by randomly assigning eyes, alternating procedures, using the same microkeratome for the initial lamellar cut, and using an adequate number of eyes in each group.
There are several limitations inherent in measuring the corneal curvature and calculating its changes utilizing a keratometer on cadaver eyes. Only two points approximately 3 mm apart are sampled, and it is difficult to obtain repeated measurements in the identical location on the cornea before and after surgery. It is possible that these keratometric measurements are taken near or on the shoulder of the ablated or resected zone - the area of the transition to unresected cornea - and thus underestimate the flattening centrally. Corneal topography would give a more complete depiction of the changes in corneal curvature.
The amount of central keratometric flattening does not correlate well with the amount of refractive change in published clinical reports of myopic keratomileusis. Consistently, there is a greater refractive reduction of myopia than keratometric flattening,8-10,12 with a ratio of refractive change to keratometric change ranging from 1.45 to 1.95.8-0,16 This difference could be due to the error induced by the keratometer in assuming the cornea to be spherical, while it is actually aspheric, or to inaccurate centering of the keratometric mires in the surgical "optical zone."16-17 Taking these factors into consideration, we may tentatively attempt to correlate our findings with clinical reports where the keratometer has also been used for measurement of corneal curvature. The excimer laser method of stromal ablation in this experiment produced an average of 5.80 D of corneal flattening. Extrapolating with an average refractive-to-keratometric change ratio of 1.63 from published reports, we could expect a reduction of myopia of 9.45 D (1.63 x 5.80 D) in patients with this technique.
A practical consideration is the handling of the corneal disc during surgery. Most excimer lasers currently used clinically are not located in the operating suite used for intraocular surgery. Hence, for excimer laser keratomileusis, the tissue must be transported to the laser site, ablated, and transported back to the operating room, while maintaining corneal disc sterility and hydration. We propose placing the corneal disc epithelial side down in the bowl of a sterile corneal preservation chamber (radius of curvature approximately 7.8 mm) and replacing the airtight cap immediately after resection. This corneal preservation chamber, with the disc inside, can be placed in a sterile container and transported to the laser room, where the laser is already set for surgery. The surgeon should wear sterile gloves, remove the corneal preservation chamber from the sterile container, and position it under the laser on a stable surface with adjustable height, such as a jack stand or an adjustable table. To minimize dehydration of the corneal disc, the cap of the preservation chamber should be removed immediately before the ablation, which should commence as soon as the laser is centered and focused on the tissue.
The gas pockets noted in three of our lasered corneal discs have not been reported in human photorefractive keratectomy or laser keratomileusis, and we think they were artifacts of previously frozen cadaver tissue due to structural changes in the corneal stroma. Histopathologic examination was not available.
The mechanical methods of tissue removal from the corneal stroma contribute several sources of error. In cryolathe keratomileusis, hydration changes, variations in tissue expansion after freezing, possible décentration of the corneal disc on the lathe, structural alteration of the lathing instrumentation at freezing temperatures, and other factors may cause significant inaccuracy. Both the nonfreeze planar and in situ methods depend on several mechanical steps where errors can be compounded. In the BKS method, error can be introduced in zeroing the lens bench, decentering the disc on the die, changes in hydration (particularly with the application of vacuum to the disc), tissue drag with the microkeratome on the second pass, and by the limitations imposed by the discreet steps between dies which do not allow a continuously variable amount of tissue resection. Problems inherent in the in situ technique include centration of the two successive keratectomies, hydration changes of the stromal bed, and limitation of chosen correction due to discreet ring height and depth plate thicknesses for the second pass. All three techniques may exhibit irregular astigmatism from nonsmooth microkeratome incisions, distortion by sutures, and variability in wound healing.
The argon fluoride (193 nm) laser removes corneal tissue with submicron accuracy, approximately 0.25 µm per pulse, and delivery systems have been designed to create a refractive profile in the ablated surface.16 Utilizing the excimer laser to resect the stromal tissue during keratomileusis both streamlines the process and reduces the sources of error. Theoretical advantages include reliable repeatability, surgical simplicity, maintenance of viable corneal tissue, and minimizing sources of error such as freezing and mechanical cutting. The refractive ablation is, therefore, not dependent on any mechanical device or surgeon factor other than appropriately aligning the ablation on the disc or stromal bed in attaining the desired reduction of corneal power. Both resection from the corneal disc or the stromal bed (in situ keratomileusis) are possible. Although the microkeratome is still used to resect the disc and its use requires considerable skill, only one pass is necessary. Improvements in microkeratome design will make this step easier and more reliable.18 The amount of correction, diameter, and depth of ablation are then selected according to the programs in the laser software, which must be refined and verified. Excimer laser keratomileusis has been done in sighted human eyes by Buratto and colleagues,19,20 and an FDA-approved, clinical study organized by Summit Technology, Ine, began in four centers in the United States in 1991.
1. Barraquer JI. Queratomileusis para la correccion de la miopia. Arch Soc Am Oftalmol Optom. 1964;5:27-48.
2. Barraquer JI, ed. Queratoplastia Refractiva. Vol 1. Bogota, Colombia: Instituto Barraquer de America; 1970.
3. Schanzlin DJ, Jester JV, Eunduck K Cryolathe corneal injury. Cornea. 1983;2:57-68.
4. Swinger CA. Lamellar refractive keratoplasty. In: Engelstein JM, ed. Cataract Surgery - Current Options and Problems. Orlando, Fla: Grane & Stratton Inc; 1984: 225-244.
5. Swinger CA, Krumeich J, Cassiday D. Planar lamellar refractive keratoplasty. Journal of Refractive Surgery. 1986;2:1724.
6. Maxwell WA. Myopic keratomileusis: initial results and myopic keratomileusis combined with other procedures. J Cataract Refract Surg. 1987;13:518-524.
7. Nordan LT, Fallor MK Myopic keratomileusis: 74 consecutive non-amblyopic cases with one year of follow-up. Journal of Refractive Surgery. 1986;2:124-128.
8. Saragoussi J, Hanna K, Jobin D, De La Messeliere S, Besson J, Pouliquen Y. Resultats du keratomileusis myopique. J Fr Ophtalmol. 1988;11:311-316.
9. Dossi F, Bosio P. Myopic keratomileusis: results with a follow-up over one year. J Cataract Refract Surg. 1987;13:417-420.
10. Swinger CA, Barraquer JI. Keratophakia and keratomileusis- clinical results. Ophthalmology. 1981;88:709-715.
11. Maxwell WA, Nordan LT. Myopic keratomileusis: early experience. Journal of Refractive Surgery. 1985;1:99-103.
12. Swinger CA, Barker BA. Prospective evaluation of myopic keratomileusis. Ophthalmology. 1984;91:785-792.
13. Barker BA, Swinger CA. Keratophakia and keratomileusis. Int Ophthalmol Clin. 1988;28:126-133.
14. Bas MA, Nano HD. In situ myopie keratomileusis initial results: part I. Refract Corneal Surg. 1991;7:223-231.
15. Duffey RJ, Vivanti NJ, Tchah H, Hoffman RF, Lindstrom RL. Paired arcuate keratotomy. Arch Ophthalmol. 1988;106:1130-1135.
16. Waring GO. Development of a syetem for excimer laser corneal surgery. Trans Am Ophthalmol Soc. 1989;88:854983.
17. Seiler T, Fantes FE, Waring GO, Hanna KD. Laser corneal surgery. In: Waring GO, ed. Refractive Keratotomy for Myopia and Astigmatism. St Louis, Mo; Moaby Year Book; 1991.
18. Hofmann R, Bechara SJ. An independent evaluation of second generation microkeratomes. Refract Corneal Surg. 1992;8:348-354.
19. Buratto L, Ferrari M. Excimer laser intrastromal keratomileusis: case reports. J Cataract Refract Surg. 1992;18:37-41.
20. Buratto L, Ferrari M, Genisi C. Myopic keratomileusis with the excimer laser: one year follow-up. Refract Corneal Surg. In this issue.
21. Nordan LT, Fallor MK Myopic keratomileusis: 74 consecutive non-amblyopic cases with 1 year of follow-up. Journal of Refractive Surgery. 1986;2:124-128.
22. Barraquer JI, Viteri E. Results of myopic keratomileusis. Journal of Refractive Surgery. 1987;3:98-101.
23. Bosc JM, Montard M, Delbosc B, Bron A, Garcher C. Nonfreeze myopic keratomileusis: retrospective study of 27 consecutive operations. J Fr Ophtalmol. 1990;13:10-16.
24. Durand L, Burillon C, Resal R. Refractive surgery and the non-freeze BKS set. Reliability of our Lyon methods. Value of our results. Bull Soc Ophtalmol Fr. 1990;90:441-443.
Summary of Published Results of Myopic Keratomileusis
Changes in Corneal Power (D) After BKS* and Excimer Laser† Myopic Keratomileusis in Human Donor Eyes