One of the most frustrating and visually disturbing complications following photorefractive keratectomy (PRK) and laser in situ keratomileusis (LASIK) is decentration of the ablation zone. Whether caused by an eccentrically displaced treatment or by intraoperative fixation error and/or drift, it can pose serious visually disabling side effects and compromised visual function, which may cause the frustrated patient and doctor to seek a predictable solution.1 Symptoms of poor visual acuity, poor vision quality, loss of contrast, halos/ glare/starburst, and multiple images (monocular polyopia) have all been reported following both PRK and LASIK in which a significant decentration has been observed.2"6
To date, the most useful imaging technique to assist in the diagnosis of clinically significant decentration has been corneal topography.7 Various methods of topographic analysis have been used to characterize and define the extent of decentration, including the center of flattest power2,8, a vector center of mass formula9, and Fourier analysis of corneal power components.10 A topographic decentration greater than 1 mm that is associated with visual symptoms typically characterizes a clinically significant decentration. The severity of symptoms may lead to a search for a possible solution in which corneal topography has historically played a major role. Various techniques including centrally placed myopic retreatment11, decentered myopic retreatment at an equal but opposite distance12*13, single incision arcuate keratotomy13, contact lens wear14, and topographically controlled retreatment15 have all been attempted, with limited success and patient dissatisfaction.
Recently, an imaging technique has been employed for objectively detennining the monochromatic aberrations of the human eye.16,17 It uses the refractive deviation of many points of light passing into the eye through the entrance pupil, and imaged as an aberration pattern on the retina. This same technology has also recently defined the aberrated wavefront pattern of normal myopic and astigmatic eyes undergoing wavefront-guided LASIK.1819 The success of this new form of aberration sensing and wavefront-guided LASIK led to our investigation of more complex and pathologic irregularities.
This study reports our first experience in diagnosing and surgically managing the complex aberrations of clinically significant decentrations using this form of wavefront sensing technology.
PATIENTS AND METHODS
Patients and Clinical Investigation
Three eyes of three patients with remarkable decentrations of the ablation zone were scheduled for vision improvement by wavefront-guided LASIK. All three patients underwent a complete ophthalmic examination including measurement of uncorrected (UCVA) and best spectacle-corrected visual acuity (BSCVA) by means of EDTRS charts (illumination 300 cd/m2), manifest refraction, corneal topography (C-scan, Technomed, Baesweiler, Germany), ophthalmometry, ultrasonic pachymetry (SP2000, Tomey, Nagoya, Japan), applanation tonometry, wavefront analysis by means of a Tscherning-type aberrometer, slit-lamp examination of the anterior segment, and contact lens ophthalmoscopy of the posterior segment. These examinations were performed preoperatively and at postoperative months 1 and 3. The demographic data and clinical histories of all three patients are listed in Table 1.
Decentration of the ablation zone was determined by obtaining the difference between preoperative and postoperative corneal topography measurements (tangential map). The area of laser ablation surrounded by a region of approximately zero power is determined by this method. The eccentricity of ablation was determined as the distance of the center of the flattened zone from the center of the entrance pupil. The center of the flattened zone was denned as the center of mass of the area in corneal topography with a refractive power of 1 diopter (D) more than the zero refractive power obtained from the topography difference map. Five corneal topographies were done preoperatively and postoperatively from each eye to achieve higher reproducibility. Statistical reproducibility to determine the center of the ablation zone with respect to the center of the entrance pupil was in the order of ±0.1 mm standard deviation.
Data for Three Eyes of Three Patients With Decentered Ablation
Patient WM had PRK with a 6-mm ablation zone in 1999. Best spectacle-corrected visual acuity before wavefront-guided LASIK was 20/25 and UCVA was 20/40. Manifest refraction was -1.75 D in sphere without a cylindrical component. The patient complained of halos and diplopia at the physiological pupil size of 4 mm in diameter, and even more so under mesopic conditions. Decentration of the ablation zone was derived from corneal topography and was in the range of 1.5 to 2.0 mm. Corneal thickness was determined by ultrasonic pachymetry, 506 µm centrally and 493 µm at the thinnest area.
Patient SU, a steroid responder, had his first PRK in 1992 and a retreatment for vision improvement in 1998. Best spectacle-corrected visual acuity and UCVA were 20/25 before wavefront-guided LASIK. The manifest refraction was difficult to predict due to the multifocal cornea; it ranged from -0.25 to +2.50 D in sphere. Again, there was no cylindrical component determined. Corneal topography yielded a decentration of the ablation zone of approximately 1.5 to 2.0 mm. The remaining central corneal thickness was 433 µm (thinnest area). The patient reported halos and ghost images, even at his natural pupil size of 3.5 mm in diameter.
Patient RE had initial LASIK in 2000 to correct her myopia of -6.50 D. Corneal topography showed a large supero-temporal decentration of approximately 2 mm at 1 month after the initial myopic correction using a 6.5-mm-diameter ablation zone. Manifest refraction was +5.00 D without a cylindrical component at 3 months after the initial treatment. BSCVA was 20/25 and UCVA was 20/200 at a physiological pupil size of 3.5 mm. The patient complained of monocular diplopia, glare, and halos.
A wavefront analyzer based on the principles of Tscherning aberrometry was used for wavefront aberration sensing. The measuring principle and technical aspects of this wavefront device are published elsewhere.16?20 The use of a frequency doubled Nd:YAG laser within the wavefront device determines the wavefront aberrations at a wavelength of 532 nm. The entrance pupils of the patients were dilated to diameters of at least 7 mm with tropicamide (Mydriaticum Dispersa, Ciba Vision, Hettlingen, Germany) during the wavefront measurements. Pupil size was measured by means of an infrared camera compatible with the wavefront measuring device. The alignment and centration with respect to the line of sight was achieved by the infrared video camera that determines the center of the entrance pupil, and a fixation target that was mounted coaxially to the optical axis of the measuring device. Five single measurements were done preoperatively and at each postoperative examination. The measurement itself took 60 milliseconds. The wavefront aberrations W(x,y) are approximated by means of Zernike polynomials, and the rmswavefront error (root mean square) including 3rd, 4th, 5th, and 6th order aberrations (27 coefficients), were used to characterize the component parts and overall wavefront aberration. All wavefront related data were calculated for a pupil diameter corresponding to the ablation zone of the wavefront-guided treatment. The Zernike coefficients were calculated according to the OSA standards ??t reporting wavefront measurements of the eye.21 The median was derived over the five independent measurements for representation of the optical aberrations. In contrast, the median of total wavefront maps (shown in the figures) was taken as a basis for the calculation of the individual ablation profiles. A measurement was considered accurate if the maximum deviation of the wavefront map height did not differ more than two times the variance.
The wavefront-guided LASIK procedures were performed under topical anesthesia (proparacaine 0.5%). In the first step, a flap with a diameter of 9.0 mm and a thickness of 130?28 ?ta was created with a superior hinge by means of the Supratome (Schwind, Kleinostheim, Germany). After photoablation, the flap was repositioned, and the interface floated with balanced salt solution. A bandage lens soaked in ofloxacine solution 0.5% was applied for the first night after surgery. Fluorometholone drops 0.1% were used two times per day for 1 week.
Wavefront sensing was possible at all examinations. No measurements were omitted due to unsuccessful wavefront sensing. The preoperative wavefront aberration maps (Figs 1C,E) of patient WM (age 25 yr) demonstrated irregularities comparable to the irregular astigmatism determined by corneal topography (Fig IA). The wavefront refraction was -1.13 D in sphere and -0.39 D, similar to the manifest refraction (Table 2). The total rms-wavefront error was 1.72 ? 0.32 ?ta and the rms-wavefront for the higher order aberrations (3rd, 4th, 5th, and 6th order) was 0.58 ? 0.11 ?ta. The Zernike expansions of the measured wavefront aberrations resulted in a spectrum (Fig IG) of extraordinary high coefficients such as horizontal tilt C2, horizontal coma C8, spherical aberration C 12, and higher order horizontal coma C18.
The optical zone for wavefront-guided LASIK was 7.0 mm, the ablation zone 9.2 mm, the central ablation depth 27 ??a, and the maximum ablation depth was 34 ?ta. (Fig 2). The operation was uneventful, as was the early postoperative follow-up.
At 3 months, BSCVA had improved from 20/25 to 20/12.5 and UCVA from 20/40 to 20/12.5. The manifest refraction was reduced to emmetropia (Table 2). Corneal topography yielded a recentered ablation zone with only small irregularities (Fig IB) within a 5-mm circle. The wavefront maps (Figs ID1F) show a decrease of the optical aberrations as determined by the decrease of the total (61%) and higher order rms-wavefront errors (33%), as shown in Table 2. In detail, the rms-values for the tilt (rmsl), sphere and cylinder (rms2), spherical-like (rms4), and 5th order aberrations (rms5) were significantly reduced. The rms for the coma-like (rms3) and 6th order aberrations (rms6) were unchanged. The spectrum of postoperative Zernike coefficients in Figure IG compared to the preoperative coefficients demonstrated a decrease, especially in horizontal and vertical tilt Cl and C2, defocus C4, astigmatism C3 and C5, and horizontal coma C8 and C 18. However, it appears that other higher order aberrations such as trefoil (C6,C9), vertical coma C7, and other higher order terms (C15, C16, C17) were slightly increased. The spherical aberration C 12 was significantly undercorrected but the 6th order spherical aberrations were found to be effectively reduced (Fig IG). The patient was free of double and ghost images but still reported glare under mesopic conditions. All reported wavefront data of patient WM were calculated for a 7.0-mm pupil diameter.
Data Before and After Wavefront-guided LASIK for Patient WM
Wavefront sensing was possible at all examinations. Two of the preoperatively performed measurements were omitted due to eye movement during wavefront sensing. Nevertheless, these two measurements were successfully repeated. The preoperative corneal topography as well as the wavefront maps of the total and higher order aberrations of patient SU are shown in Figs 3A,C,E, respectively. The wavefront refraction was -0.71 D in sphere and -0.28 D / 54? in cylinder (Table 3) and was different compared to the manifest refraction of +2.50 D in sphere. Again, the preoperative Zernike coefficients for horizontal tilt G2, horizontal coma C8, spherical aberration C12, and higher order astigmatism C13 were abnormally high (Fig 3G) when comparing published data for a normal population (Mrochen M, Kaemmerer M, Mierdel P, Krmke H-E, Seiler T. Is the human eye a perfect optic? SPIE Proceedings, Ophthalmic Technologies, San Jose, USA, 2001).
Figure 1 . Patient WM. Corneal topography: A) preoperatively, and B) postoperatively; Total wavefront map: C) preoperatively, and D) postoperatively; Wavefront map of the higher 3rd, 4th, 5th, and 6th order aberrations, including tilt: E) before wavefront-guided LASIK, and F) after wavefront-guided LASIK; (continued next page)
Figure 1... continued. G) Zernike spectrum up to the 6th order before and 3 months after wavefront-guided LASIK. The preoperative and postoperative wavefront data are reported for a 7.0-mmdiameter pupil. The preoperative rms-wavefront error of higher orders was 0.58 ? 0.11 pm and postoperatively, 0.39 ? 0.08 pm. The Zernike coefficients were reported according to OSA standards.
Visual Acuity, Refraction, and Optical Data Before and After Wavefront-guided LASIK for Patient SU
For patient SU, the optical zone was 6.0 mm, the ablation zone 8.5 mm, the central ablation depth 30 ?ta, and the maximum ablation depth was 58 ???, located in the nasal periphery of the ablation zone (Fig 4). The operation was uneventful, as was the early postoperative follow-up.
At 3 months, there was an improvement of BSCVA from 20/30 to 20/20, and UCVA improved from 20/30 to 20/25. The patient did not report halos or monocular double images. Manifest refraction was piano in sphere, whereas the wavefront refraction was -0.80 D in sphere and -0.36 D / 44? in cylinder, but was not significantly different from the preoperative wavefront refraction. Corneal topography still showed a decentered but significantly increased ablation zone (Fig 3B). The total rms-wavefront error was reduced by 28% and the higher order rmswavefront error was decreased by 21% (Table 3). This reduction in optical aberration can also be seen in the wavefront maps (Figs 3D,F) and the spectrum of Zernike coefficients shown in Figure 3G. Some aberrations such as C6, C8, ClO, and C 12 decreased 3 months after surgery; others, such as C3, C4, C7, CIl, C17, and C18 increased after surgery. All reported wavefront data of patient SU were calculated for a 6.0-mm-diameter pupil.
Wavefront-sensing failed in patient RE due to the enormous optical aberrations of higher order at the corneal surface, as shown by corneal topography in Figure 5. Thickness within the flattest area of the corneal thickness was 416 ?t?. As a consequence, a retreatment by means of wavefront-guided LASIK was not possible. This patient was informed about his situation and was scheduled for corneal transplantation.
In this study, our wavefront device, which is based on retinal imaging optics16,17,20, successfully imaged only two of three eyes with significant decentration after laser vision correction. The fact that the eye that was not imaged successfully following LASIK, rather than PRK (as the others), does not indicate that wavefront imaging of LASIK decentration is more prohibitive than PRK decentration. Rather, the postoperative LASLK eye was so severely decentered in refractive magnitude and location that wavefront imaging was not possible. It is a limitation of the wavefront measuring device that eyes with large aberrations could not be measured. Wavefront customized laser vision correction is possible only in eyes that first can be successfully imaged by a wavefront-sensing device. Eyes with aberrations too large for imaging must have another form of correction, such as lamellar or penetrating keratoplasty, as was necessary in the third patient (RE).
Figure 2. Ablation proftfes of patient WM obtained from the wavefront maps. A) Total ablation profile used for wavefront-guided LASIK; B) sphero-cylindrical treatment based on the wavefront refraction; C) ablation profile obtained from the wavefront map without the tilt and sphero-cylindrical components. The optical zone was 7.0 mm in diameter and the ablation zone 9.2 mm at maximum.
An important finding of this preliminary study was that the wavefront refraction as determined from Zernike coefficients (Eq 1) does not necessarily correlate with the manifest refraction. This certainly is true in eyes with large higher order aberrations. For example, an eye with a significant spherical aberration but without wavefront sphere at a pupil size of 6 mm can provide a significant manifest refraction (pupil size 2.5 to 4 mm) of several diopters. The main reason for this circumstance is the different pupil size, which can be simulated easily when recalculating the wavefront refraction for the same pupil size at which the manifest refraction was obtained. Consequently, the two refractions will correlate only if the subject's eye has low amounts of higher order aberrations and the wavefront measurement and manifest refraction were obtained over the same pupil size. This is important for appropriately and accurately correcting sphere and cylinder using only wavefront refraction (sphere and cylinder). From a practical point of view one should keep in mind that the manifest refraction is obtained at pupil sizes on the order of 3 mm, but the diameter of the ablation zones used in conventional corneal laser surgery is in the order of 5.5 to 7.0 mm. As a consequence, a treatment based on manifest refraction in eyes with large higher order aberrations might lead to a poor predictable result. Figure 6 shows a computer simulation performed with CTView 3.0 (Sarver and Assosiates Inc., Merrit Island, FL) for a diffraction limited eye (Fig 6A), the preoperative situation of patient WM in Fig 6B, the postoperative result of patient WM, and a simulation of a purely sphero-cylindrical correction. Using only sphero-cylindrical correction does not improve the visual acuity and might lead to an unnecessary amount of tissue removal. The simulations were performed for a 7-mm pupil.
In those eyes that were successfully examined with the wavefront-sensing device, the aberrations most notably observed were those of horizontal tilt (C2), horizontal coma (C8), and higher order astigmatism (C13). These are all asymmetric aberrations, which can be viewed as major contributors of irregular astigmatism in a horizontal decentered laser ablation. Also present among the higher order aberrations is spherical aberration (C 12), which is rotationally symmetric? also a common finding after uncomplicated conventional laser vision correction.23 The fact that each of these complex components can be measured and recorded allows us to mathematically reproduce the aberrated wavefront pattern for the computation of the ideal ablation profile and spot pattern necessary to correct this disorder.
Figure 3. Patient SU. Corneal topography: A) before wavefront-guided LASIK, and B) after wavefront-guided LASIK; Total wavefront map: C) preoperatively, and D) postoperatively; Wavefront map of the higher 3rd, 4th, 5th, and 6th order aberrations, including tilt: E) preoperatively, and F) postoperatively; continued next page)
Figure 3... continued. G) Zernike spectrum up to the 6th order before and 3 months after wavefront-guided LASIK, respectively. The preoperative and postoperative wavefront data are reported for a 7.0-mm-diameter pupil. The preoperative rms-wavefront error of higher orders was 1.00 ? 0.16 pm and postoperatively, 0.79 ? 0.12 pm. The Zernike coefficients were reported according to OSA standards.
Figure 4. Ablation profiles of patient SU obtained from the wavefront maps. A) Total ablation profile used for wavefront-guided LASIK; B) sphero-cylindrical treatment based on the wavefront refraction; C) ablation profile obtained from the wavefront map without the tilt and sphero-cylindrical components. The optical zone was 6.0 mm in diameter and the ablation zone 8.5 mm at maximum.
Figure 5. Corneal topography of patient RE. Wavefront sensing with a Tscherning-type aberrometer was not possible due to the large higher order optical aberrations in this eye.
Figure 6. Computer simulation of visual performance of patient WM. A) An ideal eye (diffraction limited); B) based on the wavefront aberrations before wavefront-guided treatment; C) demonstrates the results of wavefront-guided LASIK, and D) shows a vision chart where the calculations were based only on a pure sphero-cylindrical correction, as shown in Figure 2B. The computer simulation was done with CTView 3.0 by convolving a computer image of a vision chart with the point-spread function at a pupil size of 7.0 mm.
Figure 2 shows various types of ablation profiles that can be reconstructed on the basis of wavefront measurements. Figure 2A depicts the total ablation profile, whereas Figure 2C shows the pure spherocylindrical ablation profile obtained from the total ablation profile by removing the higher order Zernike terms. One might save almost 15 ??a in maximum ablation depth compared to the total ablation profile. On the other hand, one might use an ablation profile with a shape that is not appropriate to correct the optical errors of this eye. Figures 2A, 2B, and 2C show that customized corneal laser surgery done with a one-step procedure is, in this case, the method of choice. A multiple step procedure, such as preliminary treatment of the sphero-cylindrical errors (Fig 2C) followed by a subsequent treatment of the tilt and higher order aberration (Fig 2B), will result in an ablation depth of 48 ???; this is about 13 ??a more compared to the maximum ablation depth of the total ablation profile (Fig 2A) used for a single-step procedure. Similar conclusions can be drawn from the ablation profile (Fig 4) based on the total wavefront-map for patient SU.
So far, tilt of the wavefront has not been discussed in the literature regarding its influence in physiological optics and wavefront sensing of the human eye. Tilt characterizes a kind of prismatic effect hitherto ignored in wavefront sensing and customized ablation. Such a prismatic component in the eye should lead to a shift in the point spread function at the retinal level. In principal, the eye is able to rotate, and thus to compensate for such a prismatic error. However, the amount of prismatic error tolerated by the eye is somewhat unclear, especially in eyes with grossly decentered ablation. On the other hand, tilting the wavefront, if intentionally performed, might represent a therapeutic modality in extrafoveal fixation and age-related macular degeneration. In detail, introducing a prismatic component might shift the point spread function toward the periphery of the fovea. Further clinical studies should concentrate on the importance of prismatic errors after complicated refractive surgery and in patients with extrafoveal fixation to clarify the feasibility of such a customized treatment.
We think wavefront-guided customized laser surgery would be best performed with a one-step procedure. Customized corneal laser surgery done in a one-step procedure, as reported in our study, has potential to be an effective technique for vision improvement in patients with previously decentered ablations.
1. Azar DT, Yeh PC. Corneal topographic evaluation of decentration in photorefractive keratectomy: treatment displacement vs intraoperative drift. Am J Opthalmol 1997;124: 312-320.
2. Cavanaugh TB, Durrie DS, Riedel SM, Hunkeler JD, Lesher MP. Topographical analysis of the centration of excimer laser photorefractive keratectomy. J Cataract Refract Surg 1993;19(suppl):136-143.
3. Hersh PS, Shah SI, Geiger D, Holladay JT. Corneal optical irregularity after excimer laser photorefractive keratectomy. The Summit Photorefractive Keratectomy Topography Study Group. J Cataract Refract Surg 1996;22:197-204.
4. Verd?n W, Bullimore M, Maloney RK. Visual performance after photorefractive keratectomy. A prospective study. Arch Ophthalmol 1996;114:1465-1472.
5. Mrochen M, Eldine MS, Kaemmerer M, Seiler T, Huetz W. Improvement in photorefractive corneal laser surgery results using an active eye-tracking system. J Cataract Refract Surg 2001;27:1000-1006.
6. Mrochen M, Kaemmerer M, Mierdel P, Seiler T. Increased higher-order optical aberrations after laser refractive surgery. A problem of subclinical decentration. J Cataract Refract Surg 2001;27:362-369.
7. Wilson SE, Klyce SD, McDonald MB, Liu JC, Kaufman HE. Changes in corneal topography after excimer laser photorefractive keratectomy for myopia. Ophthalmology 1991;98:1338-1347.
8. Schwartz-Goldstein BH, Hersh PS. Corneal topography of phase DT excimer laser photorefractive keratectomy. Optical zone centration analysis. Summit Photorefractive Keratectomy Study Group. Ophthalmology 1995;102: 951-962.
9. Almendral D, Waller SG, Talamo JH. Assessment of ablation zone centration after photorefractive keratectomy using a vector center of mass formula. J Refract Surg 1996; 12: 483-491.
10. Hjortdal JO, Erdmann L, Bek T. Fourier analysis of videokeratographic data. A tool for separation of spherical, regular astigmatic and irregular astigmatic corneal power components. Ophthalmic Physiol Opt 1995;15:171-185.
11. Perez-Santonja JJ, Ayala MJ, Sakla HF, Ruiz-Moreno JM, Alio JL. Retreatment after laser in situ keratomileusis. Ophthalmology 1999;106:21-28.
12. Alkara N, Genth U, Seiler T. Diametral ablation: a technique to manage decentered photorefractive keratectomy for myopia. J Refract Surg 1999;15: 436-440.
13. Pallikaris IG, S?ganos DS1 Katsanevaki VI. LASDC complications and their management. In: Pallikaris IG, S?ganos DS, eds. LASIK. Thorofare, NJ: SLACK Ine; 1998:257-274.
14. Lim L, Siow KL, Chong JS, Tan DT. Contact lens wear after photorefractive keratectomy: comparison between rigid gas permeable and soft contact lenses. CLAO J 1999;25: 222-227.
15. Dausch D, Schroder E, Dausen S. Topography-controlled excimer laser photorefractive keratectomy. J Refract Surg 2000;16:13-22.
16. Mierdel P, Wiegend W, Krinke HE, Kaemmerer M, Seiler T. Measuring device for determining monochromatic aberrations of the human eye. Ophthalmologe 1997 ;6:441-445.
17. Krueger RR, Kaemmerer M, Mrochen M, Seiler T. Understanding refraction and accommodation through retinal imaging aberrometry. A case report. Ophthalmology 2001;108:674-678.
18. Mrochen M, Kaemmerer M, Seiler T. Wavefront-guided laser in-situ keratomileusis: early results in three eyes. J Refract Surg 2000;16:116-121.
19. Mrochen M, Kaemmerer M, Seiler T. Clinical results of wavefront-guided LASIK at 3 months after surgery. J Cataract Refract Surg 2001;27:201-207.
20. Mierdel P, Kaemmerer M, Mrochen M, Krinke H-E, Seiler T. An automated aberrometer for clinical use. SPIE Proceedings 2000;3908:86-92.
21. Thibos LN, Applegate RA1 Schwingerling JT, Webb R. Report from the VISA taskforce on standards for reporting optical aberrations of the eye. J Refract Surg 2000;16: S654-S655.
22. Klein SA. Optimal corneal ablation for eyes with arbitrary Hartmann-Shack aberrations. J Opt SoC Am A 1998;15:2580-2588.
23. Seiler T, Kaemmerer M, Mierdel P, Krinke H-E. Ocular optical aberrations after photorefractive keratectomy for myopia and myopic astigmatism. Arch Ophthalmol 2000;118:17-21.
Data for Three Eyes of Three Patients With Decentered Ablation
Data Before and After Wavefront-guided LASIK for Patient WM
Visual Acuity, Refraction, and Optical Data Before and After Wavefront-guided LASIK for Patient SU