Avoiding Serious Corneal Complications of Laser Assisted In Situ Keratomileusis and Photorefractive Keratectomy

Transcription

1 Avoiding Serious Corneal Complications of Laser Assisted In Situ Keratomileusis and Photorefractive Keratectomy Simon P. Holland, MB, FRCSC, 1,2 Sabong Srivannaboon, MD, 1,3 Dan Z. Reinstein, MD, FRCSC 1,4 Objective: To identify avoidable factors that can lead to serious complications of laser refractive surgery (photorefractive keratectomy [PRK] and laser assisted in situ keratomileusis [LASIK]). Design: Noncomparative case series. Participants: Twenty-seven eyes of 19 patients who had undergone either LASIK or PRK with severe complications accrued retrospectively. Intervention: Review of clinical records. Main Outcome Measures: Symptoms, refractive outcome, and assessment of avoidable factors contributing to the complication. Results: Patients were analyzed in four groups: group 1, scarring with ectasia; group 2, unrecognized keratoconus; group 3, flap related LASIK complications; and group 4, multiple retreatments. The 8 eyes with scarring and ectasia presented with the worst vision, 20/400 uncorrected visual acuity and 20/200 best spectacle-corrected visual acuity, with avoidable factors considered as high or difficult prescriptions with multiple retreatments. Four eyes in two patients with possible forme fruste keratoconus showed worsening irregular astigmatism. Laser assisted in situ keratomileusis flap complications included six eyes with partial laser treatment under an incomplete flap with subsequent severe irregular astigmatism. Six eyes in three patients who had undergone an average of three multiple retreatments showed decreased vision with irregular astigmatism. Conclusions: Certain severe complications of laser refractive surgery likely can be avoided by using caution when treating high prescriptions, particularly with retreatments, recognizing early keratoconus and avoiding laser treatment under a partial flap in LASIK. Ophthalmology 2000;107: by the American Academy of Ophthalmology. Originally received: November 10, Accepted: November 2, Manuscript no Department of Ophthalmology, University of British Columbia, Vancouver, British Columbia, Canada. 2 The Laser Center, Vancouver, British Columbia, Canada. 3 Department of Ophthalmology, Siriraj Hospital, Mahidol University, Bangkok, Thailand. 4 Department of Ophthalmology, Cornell University, New York, New York. Presented in part at the American Academy of Ophthalmology annual meeting, New Orleans, Louisiana, November Supported by The Eye Bank of British Columbia (SH). Correspondence to Simon P. Holland, MB, FRCSC, The Eye Care Centre, University of British Columbia, 2550 Willow Street, Vancouver, BC V5Z 3N9, Canada. Relatively few serious complications of photorefractive keratectomy (PRK) and laser assisted in situ keratomileusis (LASIK) have been reported in the peer-reviewed literature. 1 3 Most patients undergoing laser refractive surgery have corrected vision of 20/20 or close to this level. Thus any potentially vision-threatening complication is of great importance in such an elective procedure, with readily available alternatives for correcting the refractive error. We thus aimed to study patients with severe complications of laser refractive surgery referred to a single, university-based corneal refractive practice, with a goal of identifying avoidable factors that may have contributed to the complications. Methods The study was of a noncomparative case series and was performed by chart review. Institutional review board approval was not obtained. There is no established definition of a serious laser refractive complication, thus the inclusion criteria were: (1) patient referred for further opinion for diagnosis and management of a laser refractive complication; (2) loss of two or more lines of best spectacle-corrected visual acuity (BSCVA), loss of quality of vision such as to interfere with the patient s ability to function with daily activities, or both; and (3) correction of the complication likely to require further specialized laser treatment, for example, customized, corneal surgery such as penetrating keratoplasty (PKP), or both. Exclusion criteria included complications due to readily identifiable and correctable conditions such as epithelial ingrowth, incomplete LASIK flaps not treated, over- and undercorrection (enhancement) procedures, central islands, and haze amenable to further treatment. The sample was drawn from referred cases over an 18-month period from March 1, 1997 through August 31, 1998 and were accrued retrospectively. The following information was extracted for analysis from review of the patients charts: age, sex, eye, by the American Academy of Ophthalmology ISSN /00/$ see front matter Published by Elsevier Science Inc. PII S (99)

2 Holland et al Avoiding Complications of LASIK and PRK preoperative BSCVA, manifest refraction, initial procedure, laser, operative complications if any, and retreatment procedure(s). An attempt was made to obtain missing data. Data collected at the time of the consultation visit included: symptoms, uncorrected visual acuity (UCVA), BSCVA, manifest refraction, the UCVA and BSCVA in the fellow eye, corneal findings including pachymetry where this could be obtained, and diagnosis with potential management. Factors possibly affecting the development of the complication were analyzed and included: preoperative factors, high initial prescription, thin corneas, irregular astigmatism, and early keratoconus on topography. Operative complications included: flap-related complications in LASIK and -related complications in PRK. Postoperative complications included: flap complications in LASIK and scarring and ectasia in PRK. The cases were then studied to assess what was considered the most important aspect of the complication. The complications were classified into four groups: group 1, scarring, ectasia, or both; group 2, LASIK flap related; group 3, keratoconus and irregular astigmatism; and group 4, multiple retreatments. Results Twenty seven eyes of 19 patients were included in the study. Seventeen patients were operated on by one surgeon, with one patient each from two other surgeons. Four patients were referred by their refractive surgeon but most (n 14) were referred by optometrists without the knowledge of the surgeon. One patient was self-referred. Group 1 Eight patients were identified with corneal scarring, ectasia, or both, two having received PRK as the initial procedure and six having undergone LASIK. Ectasia was diagnosed by the slit-lamp appearance of corneal thinning, usually central, with central pachymetry less than 375 m when this could be recorded, and associated with topographical steepening. All patients reported decreased vision (Table 1), with variable image ghosting, doubling, and loss of depth perception. The UCVA in this group was uniformly poor at 20/400 with a range of counting fingers to 20/200. Best spectacle-corrected visual acuity was 20/200, with a range of 20/60 to 20/400. The initial refraction of the PRK patients was hyperopic astigmatism (patients 7 and 8). The initial refraction of the LASIK patients included high astigmatism (n 1), high or extreme myopia (n 4), and hyperopia (n 1). Detailed information on vision and refraction after initial treatment was available from the patients optometrists for four of the eight eyes with ectasia (patients 2, 3, and 5). These four eyes demonstrated unstable vision and refraction, with refraction changing by more than 1 diopter (D) within each 3-month period. Those eyes that underwent myopic treatment showed central corneal thinning of various degrees. Two of the three hyperopic eyes (patients 6 and 7) showed predominant thinning in the bed of the hyperopic zone, which was also adjacent to the densest area of scarring. Scarring in patient 1 was predominantly annular and associated with a double flap edge and under the hinge, from presumed epithelial downgrowth. Objective evaluation of the ectasia was difficult, because readings were often unobtainable with an ultrasound pachymeter. Orbscan (Orbtek, Salt Lake City, Utah) measurements varied between 214 m and 362 m. 4 High frequency digital ultrasound arc B scan (50 Hz) was obtained in one patient (patient 5). This patient experienced bilateral ectasia with initial refraction more than D spherical equivalent in both eyes. Patient 7 had undergone hyperopic PRK and had experienced a perforation during a hyperopic retreatment in the fellow eye, necessitating urgent penetrating keratoplasty (Holland SP, poster presentation at the AAO Annual Meeting, San Francisco, 1997). Retreatment was frequently encountered in this group, with three eyes undergoing two retreatments each and two eyes undergoing three retreatments. In the six patients in whom one eye sustained the serious complication, UCVA in the fellow eye after laser refractive treatment was better than 20/40 in only three, but all six had BCVA of 20/40 or better. Management to date has included PKP in four, and lamellar keratoplasty with homoplasty cap or PKP was recommended for the remaining patients. The left eye of the 35-year-old patient (patient 5) was illustrative of this group (Table 1). The initial refraction showed extreme myopia of D to 1.00 D 180 correcting to 20/20, undergoing LASIK with a retreatment at 1 month with flap lift and treatment of 2.25 D to 1.00 D 180. Vision at presentation was counting fingers with BCVA of 20/200 with a D. The slit-lamp examination (Fig 1) showed extreme thinning with central ectasia. The topography (Fig 2) showed a central steep area of approximately 58 D, with a surrounding area of flattening. Keratoscopy showed a double-ring image typical of localized ectasia (Fig 2). Orbscan pachymetry was 214 m, which correlated closely to the result obtained from high frequency ultrasound (Fig 3), which gave a total corneal thickness of 229 m. Figure 3 demonstrates the stromal component of the flap being 126 m, with a residual stromal bed thickness of only 46 m. Group 2 Two patients were referred with irregular astigmatism after laser refractive surgery (Table 2). Preoperative topography was obtained, and both showed inferior steepening in each eye. There are no established criteria for the topographic diagnosis of forme fruste or preclinical keratoconus, 5 8 but we considered the preoperative inferior steepening to be relevant in the patients outcome. The first patient (group 2, patient 1) was a 41-year-old female with contact lens intolerance and BCVA of 20/30, topographic power of D in the right eye and D in the left eye. The patient underwent bilateral PRK, apparently centered on the visual axis, with immediate postoperative image doubling and ghosting in each eye. A retreatment was performed, with no improvement in symptoms. The pre- and postoperative topography of the left eye is shown in Figure 4. Uncorrected visual acuity was 20/30 in both eyes, with the patient deemed to be incapacitated as a result of the poor quality of vision. Central pachymetry was 375 and 381 m; preoperative pachymetry was not available. The patient was not informed of the possibility of having keratoconus before surgery. The second patient was also contact lens intolerant (rigid gas permeable). Inferior steepening was noted, but the superior inferior difference was only 2 D at 3 mm(fig 5). Topographical power was D in the right eye and D in the left eye. The patient underwent bilateral LASIK 6 weeks after discontinuing contact lens wear. The patient was told that the stability of the refraction was questionable and that the best visual expectation would be 20/30. Preoperative pachymetry was 510 m in the right eye and 516 m in the left eye. The patient underwent bilateral myopic/ astigmatic LASIK, and image doubling developed in both eyes. Vision, however, improved to 20/20 in both eyes, but she remained highly symptomatic. The irregular astigmatism on topography increased to nearly 10 D in the right eye and 6 D in the left eye, although there was no increase in the inferior steepening [Figure 6]. The patient returned 15 months later and underwent a custom- 641

3 Ophthalmology Volume 107, Number 4, April 2000 Table 1. Group 1, Scarring and Patient 1 40-Year-Old Male, Right Eye Patient 2 26-Year-Old Female, Patient 3 46-Year-Old Male, Right Eye Patient 4 35-Year-Old Male, Preoperative refraction 1.25 to to / to to /50 Initial procedure LASIK LASIK LASIK LASIK Complication Incomplete flap & immediately recut 2 nd flap No No No Retreatment Presenting symptoms & signs 1 month, flap lift with myopic 4.50 to /30 Ghost images, decreased depth perception Examination UCVA 20/400 Refraction: 1.00 to Dense annular scar with central haze and area of epithelial ingrowth, double edge of flap, peripheral thinning Topographic power (D) Central corneal thickness ( m) Fellow eye UCVA 20/ to after uncomplicated LASIK 1) 2 months, flap lift with hyperopic 5.75 to /30 2) 3 months, flap lift with myopic astigmatism plano /70 Decreased vision, double image UCVA 20/400 Refraction: to Central corneal thinning with ectasia Three retreatments, at 1 month, 3 months, and 1 year after operation; no further information Multiple and shadowing of images UCVA 20/400 Refraction: 3.00 sph 20/400 Central corneal thinning with ectasia, multiple linear scars underneath flap ) 6 months, flap lift with myopic astigmatism 2.50 to /40 2) 1 year, flap lift for epithelial ingrowth Decreased vision UCVA 20/400 Refraction: 1.75 to \1 20/400 Central corneal thinning with ectasia, short flap N/A 273 N/A Nonrecordable by pachymeter UCVA 20/30 UCVA 20/20 UCVA 20/50 Plano sph 1.25 to /25 after after myopic 20/40 after uncomplicated LASIK myopic LASIK LASIK Management PKP PKP Consider PKP/lamellar keratoplasty Final visual acuity 1.50 to 2/ to /25 Consider PKP/lamellar keratoplasty LASIK laser assisted in situ keratomileusis; N/A not available; PKP penetrating keratoplasty; PTK phototherapeutic keratectomy; RGP rigid ized with an optical zone of 4 mm, decentered 2 mm inferiorly. Further follow-up is unavailable. Group 3 Six patients involving a total of 8 eyes sustained flap-related complications (Table 3). Four patients had low myopia, having an average refraction of 4.50 D (range, 3.50 to 4.75), one patient had high hyperopia, ( 5.25/ 0.50 D 180), and one patient had high myopia ( 10.50/ 0.50 D 180). An incomplete or short flap was obtained in 7 eyes. One was treated the following day with a PRK through the flap, and the other six received laser on part of the intended stromal bed at the time of the surgery. The eighth eye, the fellow eye of the one treated with PRK on the first postoperative day, was left with an oblique fold across the visual axis. One patient underwent retreatment with a recut 1 month after the initial procedure, with the fellow eye undergoing a flap lift for epithelial ingrowth. Otherwise, none of the patients have so far received retreatments. The six patients reported monocular diplopia of varying degrees with distorted and decreased vision, especially at night. On examination, the presenting UCVA ranged between 20/25 and 20/200. Five of the eight eyes achieved BSCVA of 20/40 or better, but were unable to tolerate the refraction because of the image distortion and ghosting. Topographic features were of highly irregular astigmatism with a steep nasal cornea consistent with a partial of the stromal bed. Initial management was rigid contact lenses when these could be tolerated, with the possibility of further laser treatment and customized, such as topographically assisted. 9 Figure 7 shows the left cornea of patient 642

4 Ectasia after Laser Refractive Surgery Holland et al Avoiding Complications of LASIK and PRK Patient 5 35-Year-Old Male Patient 6 62-Year-Old Male, Right Eye to to to Patient 7 52-Year-Old Male, Right Eye 3.75 to LASIK LASIK LASIK PRK PRK No No No No No Patient 8 57-Year-Old Male, Right Eye 6.75 to /25 1 month, flap lift with myopic 2.25 to /20 1 month, flap lift with myopic 2.25 to /20 1) 3 months, flap lift with laser 1.00 to /80 2) 8 months, myopic surface 1.00 sph 20/400 1) 1 month, hyperopic PRK enhancement 4.25 to /40 2) 2 months, hyperopic PRK enhancement 5.50 to /60 No Decreased vision Decreased vision Decreased vision Decreased vision Decreased vision UCVA 20/200 Refraction: 1.25 to /60 Central corneal thinning UCVA counting fingers Refraction: /200 Severe central corneal thinning with ectasia UCVA 20/400 Refraction: to / 70 Short, thin flap with interface haze UCVA 20/400 Refraction: no improvement Scarring and thinning UCVA 20/200 Refraction: no improvement Elevated central scar with haze and thinning No further information UCVA 20/ to /25 post myopic LASIK PKP for perforation UCVA 20/ /25 after PRK Consider PKP/lamellar keratoplasty Consider PKP/lamellar keratoplasty Consider PKP/lamellar homoplastic graft PKP 3.00 to /25 Consider lamellar homoplastic graft gas permeable contact lens; UCVA uncorrected visual acuity. 2 with a short flap, and Figure 8 shows corneal topography of the same cornea. Group 4 Complications developed in six eyes of three patients after multiple retreatments (Table 4). Multiple retreatments included preoperative refraction for one patient with extreme myopia, for whom spherical equivalent was D in the right eye and D in the left eye. The second patient had low to moderate myopia: 4.25 D in the right eye and 5.25 D in the left eye. Preoperative refraction for the third patient was not available. The patient with extreme myopia underwent initial PRK with subsequent hyperopic enhancements, followed by a laser thermal keratoplasty (LTK). The patient with low myopia underwent three enhancement procedures in one eye and four in the other, with one LTK enhancement. The LASIK patient underwent two retreatments in each eye, with an additional PRK enhancement on one eye. Symptoms were decreased vision, image doubling, halos, and headaches. Findings on examination showed a wide range of UCVA, from 20/25 to 20/400, with irregular astigmatism on topography. The fellow eye of the LASIK patient went on to have central thinning with high hyperopia of to 3.25 D 50. Advised management ranged from penetrating keratoplasty; lamellar keratoplasty with homoplasty cap; and customized, such as decentered phototherapeutic keratectomy, or topographical- 9 or ultrasound-linked. The total number of retreatments performed was 17 for the 6 eyes, so that each received an average of 3 retreatments. Figure 9 shows corneal topography of the right cornea of patient 1 in this group and shows a flat central cornea 643

5 Ophthalmology Volume 107, Number 4, April 2000 Figure 1. Slit-lamp examination of the left cornea in patient 5 (group 1) showing central ectasia of 229 m. Figure 2. Corneal topography and keratoscopic mires (Eyemap, Alcon Surgical, Fort Worth, Texas) of the same cornea in Figure 1, shows the severe irregularity and double-ring image seen in ectasia. with steepening superiorly, with irregular flattening along the LTK scars inferiorly resulting in the irregular astigmatism. Discussion Laser refractive surgery, particularly LASIK, is expanding rapidly and even a low rate of serious complications is of great importance. We studied 27 eyes of 19 patients (PRK, n 8; LASIK, n 19) to determine if there were avoidable factors in the development of the complication. Patient were considered to have serious complication if they were visually disabled to the extent they had difficulty with their daily activities, either by virtue of reduced BCVA of more than two lines or decreased quality of vision, with symptoms such as monocular diplopia. The patients who were included had complications for which solutions were not readily available, and thus needed to be considered for corneal surgery such as PKP or lamellar keratoplasty with homoplasty cap, customized laser treatment, or both. The 644

6 Holland et al Avoiding Complications of LASIK and PRK Figure 3. High frequency digital ultrasound arc B scan (50 MHz) of a 9-mm diameter horizontal section through the left cornea of patient 5 (group 1). The surface of the epithelium, Bowman s layer, and the lamellar interface are labeled. Examination of the lamellar interface contour demonstrates sudden steps peripherally depicting the outer limits of the zone within the 6.5-mm treatment zone that can be appreciated by cross-referencing to the inner corneal surface contour. Centrally, the epithelial thickness was 57 m, the stromal component of the flap was 126 m, and the residual stromal bed thickness 46 was m. Note that the lamellar interface was invisible at the slit-lamp (Fig 1), but it clearly visualized on ultrasound scanning. study has an inherent bias, because it was drawn from the referral base of a single, university-based corneal refractive practice. Only four patients were directly referred by their refractive surgeon. Most were referred by optometrists without the knowledge of the surgeon, in most cases because the patient had become dissatisfied with the progress in managing the complication. This created additional problems in management, because preoperative and operative informa- Preop refraction Table 2. Group 2, Unrecognized Keratoconus and Irregular Astigmatism Patient 1, 41-Year-Old Female Patient 2, 46-Year-Old Female Right Eye Right Eye 4.00 to /30 K 50/45.8@ to /30 K 50.5/46.1@ to /30 K 46.8/42.8@6 Initial procedure PRK PRK LASIK LASIK Complication No No No No Retreatment Retreatment of 0.75 to Retreatment of Plano No No Presenting symptoms & signs Examination Topographic power (D) Central corneal thickness ( m) Management Monocular diplopia Monocular diplopia Blurred and double vision Refraction: 2.00 to /30 Clear cornea with inferior steepening and irregular astigmatism of topography Refraction: 0.25 to /30 Clear cornea with inferior steepening and irregular astigmatism of topography Refraction: 0.75 to /20 Haze 1 at interface and inferior steepening on topography PTK Customized Consider RGP, PTK Customized Enhancement LASIK with customized from topography 3.50 to /40 K46.3/42.3@164 Blurred and double vision Refraction: 1.50 to /20 Clear cornea inferior steepening on topography Enhancement LASIK with customized from topography LASIK laser in situ keratomileusis; RGP rigid gas permeable contact lens; PTK phototherapeutic keratectomy. 645

7 Ophthalmology Volume 107, Number 4, April 2000 Patient 1 34-Year-Old Male, Right Eye Patient 2 36-Year-Old Female, Table 3. Group 3, Flap- Patient 3 31-Year-Old Female, Preop refraction 5.25 to / to / Initial procedure LASIK LASIK LASIK Complication Short flap with laser, difficult to obtain suction Short flap with laser Short flap with laser Retreatment No No No Presenting symptoms & signs Monocular diplopia, poor night vision Multiple images, decreased vision Decreased vision and image distortion Examination UCVA 20/200 Refraction: 2.50 to (20/25) Short flap with inferior deep oblique folds UCVA 20/100 Refraction: 1.00 to /40 Short flap, irregular astigmatism on topography UCVA 20/100 Refraction: no improvement Short flap across visual axis Topographic power (D) Fellow eye UCVA 20/ to /15 after hyperopic LASIK UCVA 20/ to /20 after myopic LASIK UCVA 20/ after myopic LASIK Management Consider topography-linked Consider topography-linked Consider topo-linked LASIK laser in situ keratomileusis; PRK photorefractive keratectomy; RGP rigid gas permeable contact lens; UCVA uncorrected visual acuity. tion, which could be of help to further management, was unavailable. Our study involved 27 case reports, but we found evaluating these in groups was helpful to consider avoidable factors of these serious complications. All patients had irregular astigmatism sufficient to cause significant symptoms. Group 1 In the first group of patients, characterized by corneal scarring, ectasia, or both, eight of the nine eyes had extremely poor presenting vision with average UCVA of 20/400 and BSCVA of 20/200. Two eyes (patients 1 and 8) had decreased vision primarily from corneal scarring with peripheral thinning, with six eyes showing central ectasia. Factors that may have contributed to developing scarring with ectasia included high preoperative refraction (high astigmatism, extreme hyperopia, and high and extreme myopia) and multiple retreatments. The average eye underwent two retreatments. There was insufficient information on the preoperative pachymetry to draw any conclusion, although thin corneas with high corrections are regarded as risk factors for corneal ectasia. 10 There is increasing concern regarding the occurrence of keratectasia after LASIK, as reported in several recent case reports, but the cause and mechanism remain unknown. 15 Seiler and Quurke 11 has suggested that the anterior flap after LASIK does not contribute to the biomechanics of the cornea, and thus the residual stromal bed is the critical factor in corneal stability after lamellar surgery. Machat 10 has suggested that in LASIK, a minimal stromal depth of 200 m and preferably 250 m be left, or the 50% rule with a limit of total flap thickness of no more than 50% of the total corneal thickness. However, Seiler et al 12 also quotes Barraquer, 16 who recommended a minimal thickness of 300 m of posterior stromal bed. Presently, we have no readily available method to test posterior corneal thickness. Koch 15 noted that in the five reported cases of keratectasia thus far after LASIK (in the English language), the residual stromal depth had to be calculated using multiple assumptions. Three of the four cases described by Seiler et al 11,12 had stromal bed thickness of fewer than 250 m, with the case with 265 m considered to be forme fruste keratoconus. Of greater concern is the report by Geggel and Talley, 13 who described a patient who had only 6.6 D of correction in two surgeries, with estimated posterior corneal thickness of 289 m, assuming a flap thickness of 160 m. This was likely because the subsequent microscopy of the recipient button showed that the flap was not excessively thick. Keratectasia appears to have developed in this case without evidence of preoperative forme fruste keratoconus, a high correction, or an unusually thin cornea. Our data includes five eyes with keratectasia after LASIK, and is limited by lack of preoperative information, particular pachymetry, but several factors may be relevant. All eyes (Table 1) had high or extreme myopia, with three having spherical equivalents of D or more. All were retreated an average of twice in each eye. One of the 646

8 Holland et al Avoiding Complications of LASIK and PRK Related Complications Patient 4, 47-Year-Old Female Patient 5, 35-Year-Old Male Patient 6 36-Year-Old Female, Right Eye Right Eye Approximately 3.50 Approximately LASIK LASIK LASIK LASIK LASIK None noted Incomplete flap; aborted procedure Short flap extended with a blade and Short flap extended with a blade and Short flap with laser No Double images UCVA 20/25 Refraction: /25 Oblique fold in flap with interface debris 1 day, retreatment with PRK through flap Decreased vision with double images UCVA 20/50 Refraction: Plano /50 Short flap with central haze and decentered on topography treated with laser 1 month, retreatment (recut) Night diplopia, ghost images and halos UCVA 20/70 Refraction: 1.00 to /40 Short flap with interface particles Topographically irregular astigmatism treated with laser 1 month, flap lift for epithelial ingrowth Jumbled images at night UCVA 20/50 Refraction: plano /50 Short flap with epithelial ingrowth at the hinge-edge Topographically irregular astigmatism No Headaches with decreased vision and image distortion UCVA 20/40 Refraction: 1.00 to /25 diplopia Short flap Topographically irregular astigmatism UCVA 20/ after myopic LASIK Consider flap lift and attempt to remove fold Consider RGP Contact lens Consider ultrasoundlinked Consider ultrasoundlinked Consider topographylinked ed difficulties experienced in this study was accurate depth measurement of a scarred ectatic cornea, because ultrasound pachymetry often did not give readings. Optical pachymeters such as the Orbscan and Par systems (Par Technology, New Hartford, NY.) may allow readings in these circumstances, but the accuracy has not been well evaluated. 4,17 The most valuable data on post LASIK corneal anatomy came from the use of very high frequency ultrasound (Fig 3). 18 This was able to show a residual stromal bed thickness of 46 m in the left eye of a patient with extreme myopia, who had undergone one myopic enhancement, subsequently experiencing severe corneal ectasia. This is not yet commercially available, but appears likely to be one of the more valuable methods of determining residual stromal bed and total corneal thickness before performing retreatments. 18 Thus we are unable to draw conclusions based on these six cases of keratectasia (five after LASIK and one after PRK). As far as we know, no cases developed after the primary treatment. It seems probable that performing a peripheral with a hyperopic retreatment after a deep myopic, as in at least two of the eyes, would increase the likelihood of biomechanical weakening of the cornea, rather than one localized area at the center of the myopic zone. We can only suggest that extreme caution be used in treating eyes with high or extreme corrections, always calculate the preoperative pachymetry even in low myopias, and, based on present knowledge, attempt to leave 250 m of stromal bed. However, there may be another set of adverse outcomes for extreme prescriptions, if corneas at risk for keratectasia instead undergo PRK or a small zone is used or a thinner keratome plate, such as 130 m to conserve the stromal bed. Thus we would recommend that refractive surgeons recognize the potentially increased risks of patients with high or unusual prescriptions, be cautious regarding retreatment, and report all cases of ectasia after laser refractive surgery as suggested by Koch. 15 Group 2 Two patients experienced monocular diplopia in both eyes attributed to irregular astigmatism, with one having undergone PRK and the other LASIK. To our knowledge, there is no agreed upon system to separate patients with preclinical or forme fruste keratoconus from those with irregular astigmatism. Indices have been developed by Rabinowitz and McDonnell 6 and are available based on discriminant analysis as described by Maeda et al. 5 Such methods show high sensitivity and specificity in the ability to detect clinical keratoconus. The question arises as to whether eyes with irregular astigmatism or forme fruste keratoconus respond differently from those without. Our second patient would not generally meet the criteria for keratoconus, in that the inferior steepening was mild (46.6 D) and topographic power was less than 45 D. The preoperative BCVA was, 647

9 Figure 4. Pre- and postoperative topography of the left eye of patient 1 (group 2) showing worsening irregular astigmatism. Figure 5. Preoperative corneal topography (Humphrey instrument, Zeiss Humphrey Systems, Dublin California) in patient 2 (group 2), showing inferior steepening. Figure 6. Postoperative corneal topography of the same patient in Fig 5 with increase in irregular astigmatism from preoperative topography.

10 Figure 7. Slit-lamp examination of the left cornea patient 2 (group 3) with a short flap showing the flap hinge at the edge of the pupil. Figure 8. Corneal topography and keratoscopic mires (Eyemap, Alcon Surgical, Fort Worth, Texas) of the same cornea in Figure 7 showing irregularity over the visual axis. There is an abrupt change of corneal power between the ablated and nonablated zone, separated by the line representing the hinge of the flap. Figure 9. Corneal topography of the right cornea of patient 1 (group 4) shows a flat central cornea with steepening superiorly and irregular flattening along the laser thermal keratoplasty scars inferiorly, resulting in the irregular astigmatism. 649

11 Table 4. Group 4, Multiple Retreatment Patient 1, 41-Year-Old Female Patient 2, 29-Year-Old Male Patient 3, 39-Year-Old Male Right Eye Right Eye Right Eye Preop to to N/A N/A refraction Initial PRK PRK PRK PRK LASIK LASIK procedure Complication No No No No N/A N/A Retreatment 1) PRK hyperopic enhancement 2) LTK enhancement No further information 1) PRK hyperopic enhancement 2) LTK enhancement No further information Three PRK enhancements No further information Two flap lifts with retreatment No further information Presenting symptoms & signs Four PRK hyperopic enhancements and LTK enhancement No further information Decreased vision Decreased vision Double images Halos and decreased vision Examination UCVA 20/400 Refraction: 5.00 to /60 Central corneal scarring with thinning Topographically irregular astigmatism Topographic power (D) Central corneal thickness ( m) Management UCVA counting finger Refraction: 6.25 to /50 Central corneal scarring with thinning Topographically irregular astigmatism UCVA 20/25 Refraction: /20 Clear cornea with topographically central small treatment zone with nasal steepening UCVA 20/200 Refraction: 1.25 to /30 LTK scars with topographically irregular astigmatism with inferotemporal steepening Decreased vision UCVA 20/50 Refraction: 0.75 to /25 Peripheral subepithelial scarring underneath flap with multiple interface particles N/A N/A N/A N/A N/A N/A Consider lamellar homoplastic graft Ophthalmology Volume 107, Number 4, April 2000 Consider lamellar homoplastic graft Consider topolinked Consider topolinked Consider lamellar homoplastic graft Two flap lifts with retreatment One PRK enhancement No further information Dizziness and headaches with decreased vision UCVA 20/200 Refraction: to /60 Central ectasia with peripheral scarring and interface particles Consider PRK LASIK laser in situ keratomileusis; LTK laser thermal keratoplasty; N/A not available; PRK photorefractive keratectomy. however, decreased at 20/30 in both eyes. The patient was highly symptomatic after surgery and had no change in inferior keratometry, but had flattening superiorly, and thus increase in the irregular astigmatism. It is possible that this patient had contact lens-induced corneal warpage, because refractive and topographic stability were not established before surgery. The patient who underwent PRK (patient 1) had preoperative features more suggestive of forme fruste keratoconus, with topographic power more than 48 D, also with reduced BCVA. This patient also responded with worsening irregular astigmatism. Demirbas and Pflugfelder 8 have suggested the value of elevation mapping to locate the corneal apex, which may improve on the inaccuracies of using a placido-based topographical system and may clarify concepts such as the displaced apex syndrome. 19 Ultrasound pachymetry may not be reliable to diagnose keratoconus because of higher false-negative and false-positive rates than those obtained by video keratography. 20 All four eyes in our series showed worsening symptoms and irregular astigmatism after axis laser treatment. However, Mortensen et al 21 have described a success rate of 58% in 23 patients undergoing PRK to reduce the steepness of the cone in keratoconus. Seiler and Quurke 11 have reported iatrogenic ectasia after LASIK in forme fruste keratoconus, despite residual stromal bed thickness of 265 m. 11 Thus a cornea with keratoconus-like features may be more likely to show increased elasticity. 22 The issue of safe laser refractive treatment of irregular astigmatism and forme fruste keratoconus remains unclear. Until there is further information, we would suggest using caution in treating such patients and consider factors such as topographic power, thin corneas, topographical superior/inferior disparity, and decreased BCVA before treating. Group 3 The third group involved patients who experienced flaprelated complications during LASIK. Farah et al 23 report a cumulative percentage of 2.5% of incomplete cuts based on three studies, and 3% of short flaps based on one study. The 650

12 Holland et al Avoiding Complications of LASIK and PRK recommended management after creating a short or incomplete flap is not to attempt treatment or manual extension, but to replace the flap, wait approximately 3 months, and recut a new flap. 24 In our study, the flap was extended with a blade in one case, and six of the eight eyes underwent laser treatment of the partially exposed stromal bed. One eye underwent PRK on the surface of the incomplete flap the following day. In the six patients in whom an incomplete zone was achieved under a short flap, all experienced highly irregular astigmatism and were markedly symptomatic. The eye that was treated with PRK on the first postoperative day on a short flap with experienced haze, had decentered, and experienced visually significant haze. The fellow eye of this patient had an oblique fold across the visual axis, with subsequent diplopia. These groups of patients have also proved difficult to treat, because contact lens fitting has been problematic as a result of the highly irregular astigmatism. Further laser treatment is likely to be difficult, possibly requiring topographically or ultrasound-assisted treatment. The avoidable factor in this group of patients with short or incomplete flaps is to avoid proceeding with laser on a partially exposed stromal bed. Allowing an incomplete flap to heal and performing a recut after several months with of the planned treatment zone would be likely to give an excellent prognosis. Also, because two patients in the study had bilateral flap complications, it may be advisable not to proceed with the second eye, should a problem develop with the first during bilateral procedures. Group 4 The fourth group of patients included three who underwent multiple retreatments in both eyes. The 6 eyes in this study underwent a total of 17 retreatments, that is, approximately 3 retreatments for each eye. One eye went on to scarring with ectasia and thus also could be included in the first group. The fellow eye of this patient now shows residual astigmatism after two retreatments. The left eye of the second patient with low myopia subsequently underwent four hyperopic PRK retreatments and one LTK retreatment, resulting in UCVA of 20/200 with BSCVA of 20/30. It is difficult to speculate on the avoidable factors, except that in the four eyes where the preoperative refraction was known, they were myopic and all four eyes were treated with hyperopic enhancements, and three were treated with LTK (Table 4). Thus it may be advisable to wait at least 3 months for a stable refraction before considering retreatment for hyperopic over correction with myopic PRK. It may also be advisable, pending further information, to be cautious in considering LTK after hyperopic PRK retreatment for myopia overcorrection. Conclusion We are unable to draw definite conclusions as to the avoidable factors for serious complications of LASIK and PRK based on this study. Despite the inherent biases in the case selection and definitions, we are concerned by the number of serious complications: 27 eyes in 19 patients accumulated in 18 months. Recognizing the difficulties in drawing conclusions from this group of patients, we do consider that there are several avoidable factors that other laser refractive surgeons may consider in preventing similar problems. This is especially relevant, given the poor refractive outcome of most of the patients, particularly those with scarring and ectasia with presenting UCVA of 20/400 and BSCVA of 20/200. Serious complications such as scarring with ectasia are more likely to occur in eyes with high prescriptions and undergoing multiple retreatments, particularly hyperopic retreatment after initial myopic or astigmatic PRK or LASIK. The poor results of patients who underwent partial treatment under an incomplete or short LASIK flap support the view that the preferred management of a short flap complication is not to treat with the laser at the time, nor attempt extension manually. Also in planned bilateral surgery, a flap complication in the first eye should preclude treatment of the second eye. Patients showing inferior steepening on topography, particularly if associated with steep keratometry, should be approached cautiously with laser refractive surgery. Although other authors reported success with laser treatment for keratoconus, this may require more customized, such as performing a spherical treatment on the steeper area of the cornea and careful patient counseling. 17 Caution is also advisable before performing a third or fourth retreatment for under- or overcorrection. References 1. Loewenstein A, Lipshitz I, Varssano D, Lazar M. Complications of excimer laser photorefractive keratectomy for myopia. J Cataract Refract Surg 1997;23: Alio JL, Artola A, Claramonte PJ, et al. Complications of photorefractive keratectomy for myopia: two year follow-up of 3000 cases. J Cataract Refract Surg 1998;24: Azar DT, Farah SG. Laser in situ keratomileusis versus photorefractive keratectomy: an update on indications and safety [editorial]. Ophthalmology 1998;105: Yaylali V, Kaufman SC, Thompson HW. Corneal thickness measurements with Orbscan Topography System and ultrasonic pachymetry. J Cataract Refract Surg 1997;23: Maeda N, Klyce SD, Smolek MK, Thompson HW. Automated keratoconus screening with corneal topography analysis. Invest Ophthalmol Vis Sci 1994;35: Rabinowitz YS, McDonnell PJ. Computer-assisted corneal topography in keratoconus. Refract Corneal Surg 1989; 5: Kalin NS, Maeda N, Klyce SD, et al. Automated topographic screening for keratoconus in refractive surgery candidates. CLAO J 1996;22: Demirbas NH, Pflugfelder SC. Topographic pattern and apex location of keratoconus on elevation topography maps. Cornea 1998;17: Wiesinger-Jendritza B, Knorz MC, Hugger P, Liermann A. Laser in situ keratomileusis assisted by corneal topography. J Cataract Refract Surg 1998;24: Machat JJ. Excimer Laser Refractive Surgery. Practice and Principles. Thorofare, NJ: SLACK Inc., 1996;

13 Ophthalmology Volume 107, Number 4, April Seiler T, Quurke AW. Iatrogenic keratectasia after LASIK in a case of forme fruste keratoconus. J Cataract Refract Surg 1998;24: Seiler T, Koufala K, Richter G. Iatrogenic keratectasia after laser in situ keratomileusis. J Refract Surg 1998;14: Geggel HS, Talley AR. Delayed onset keratectasias following laser in situ keratomileusis. J Cataract Refract Surg 1999; 25: Speicher L, Göttinger W. Progressive keratektasie nach laserin-situ keratomileusis (LASIK) [Eng. abstr]. Klin Monatsbl Augenheilkd 1998;213: Koch DD. The riddle of iatrogenic keratectasia [editorial]. J Cataract Refract Surg 1999;25: Barraquer JI. Queratomileusis y Queratofaquia. Bogota: Instituto Barraquer de America, 1980; Schultze RL. Accuracy of corneal elevation with four corneal topography systems. J Refract Surg 1998;14: Reinstein DZ, Silverman RH, Sutton HFS, Coleman DJ. Very high-frequency ultrasound corneal analysis identifies anatomic correlates of optical complications of lamellar refractive surgery. Anatomic diagnosis in lamellar surgery. Ophthalmology 1999;106: Doyle SJ, Hynes E, Naroo S, Shah S. PRK in patients with a keratoconic topography picture. The concept of a physiological displaced apex syndrome. Br J Ophthalmol 1996;80: Rabinowitz YS, Rasheed K, Yang H, Elashoff J. Accuracy of ultrasonic pachymetry and videokeratography in detecting keratoconus. J Cataract Refract Surg 1998;24: Mortensen J, Carlsson K, Öhrström A. Excimer laser surgery for keratoconus. J Cataract Refract Surg 1998;24: Andreassen TT, Simonsen AH, Oxlund H. Biomechanical properties of keratoconus and normal corneas. Exp Eye Res 1980;31: Farah SG, Azar DT, Gurdal C, Wong J. Laser in situ keratomileusis: Literature review of a developing technique J Cataract & Refractive Surg 1998;24: Slade SG. LASIK complications. In: Machat JJ, ed. Excimer Laser Refractive Surgery. Practice and Principles. Thorofare, NJ: SLACK Inc., 1996; HSK A beautiful virus named Jan Worked out a wonderful plan. Her figure dendritic Proved Viropto-lytic, And Metaherpetic she d play with her man. William C. Conrad, M.D. Warner Robins, GA 652