LASIK performed using a mechanical microkeratome has been related to complications affecting quality of vision, mainly due to the induction of astigmatism, increased corneal aberrations, and other effects of a nonplanar flap.1-9 The main source of undesirable functional effects and complications of a mechanical microkeratome is related to either the imprecision and progressive deterioration of mechanical elements or to the compression on the corneal surface.10
The femtosecond laser, introduced into refractive surgery in 2002, uses microphotodisruptive pulses that are focused onto a precise plane of the cornea. Thus, the lamellar cut is smoother, and the anterior and posterior faces are parallel.9,11-16 This system of cutting is not only more accurate than the microkeratome, but it is also safer, dramatically reducing the rate of potential severe complications (eg, buttonholes, epithelial defects, free caps, irregular flaps due to suction loss). As with any new technology, however, new complications and incidences have appeared. Our investigational team has studied the new inflammatory events specifically related to the use of the femtosecond laser. These are transient light sensitivity syndrome (TLSS) and energy-related diffuse lamellar keratitis (DLK).
TRANSIENT LIGHT SENSITIVITY SYNDROME
The term TLSS has been used to describe a clinical condition characterized by unusual photosensitivity; normal visual acuity returns several weeks after otherwise uneventful LASIK with the femtosecond laser. Typically, TLSS responds to topical steroid treatment.17 Although the etiology of TLSS is unknown, clinical clues point toward an inflammatory origin. Proposed mechanisms include inflammation caused by necrotic cell debris or subproducts of gas bubbles, cytokines migrating from the flap interface to the perilimbal sclera and iris base, or activated keratocytes in the interface.17
In order to describe the incidence of TLSS and possible preventive strategies, we conducted a study including the first 765 consecutive eyes we operated on with the femtosecond laser18 between July 2004 and September 2005. Patients were older than 18 years of age and had a stable refractive history correctable by LASIK. Exclusion criteria included BSCVA worse than 20/80, scotopic pupil diameter larger than 7 mm, cataract, history of uveitis or retinal detachment, corneal dystrophy, and glaucoma. The preoperative spherical equivalent (SE) ranged from -11.50 D to 6.00 D.
After the usual LASIK preoperative assessment, Gonzalo Muñoz, MD, PhD, performed all surgeries using the 15-KHz IntraLase femtosecond laser (Advanced Medical Optics, Inc., Santa Ana, California) and the Visx Star S2 Excimer Laser (Advanced Medical Optics, Inc.) at 10 Hz with a 6.5-mm optical zone. The initial settings were an attempted flap diameter of 9.2 mm, flap thickness of 110 mm, pocket width of 2.6 mm with spot separation of 6 µm, hinge angle of 50º, raster energy of 1.8 mm with interbeam separation of 11 mm, sidecut angle of 70º, and sidecut energy of 2.3 µJ.
The pattern of bubbles after femtosecond laser ablation was classified according to the presence/absence of an opacified bubble layer. Immediately after the ablation with the femtosecond laser, the surgical bed was rotated, and the patient was positioned under the excimer laser, minimizing the time allowed for absorption of the microcavitation bubbles.
In the first 287 eyes (153 patients), a low-dose regimen of topical steroids consisting of tobramycin and dexamethasone eyedrops (Tobradex; Alcon Laboratories, Inc., Fort Worth, Texas) was used three times daily for 7 days. The following 478 eyes (261 patients) received a high-dose regimen of topical steroids consisting of Tobradex eight times daily for the first 3 days and then four times daily for 1 week. Lubricating eyedrops (Acuolens; Alcon Cusi, Barcelona, Spain) were used three to four times daily for 3 months.
Patients were examined 1 day, 1 week, and 1 and 6 months postoperatively. Data recorded during postoperative visits included UCVA, BSCVA, slit-lamp examination, applanation tonometry, and corneal topography.
RESULTS
Of the 414 patients who underwent LASIK with the femtosecond laser, five reported severe light sensitivity in both eyes. Both eyes in two patients underwent hyperopic LASIK, and both eyes in three patients underwent myopic LASIK. Photophobia presented within 6 to 8 weeks after LASIK and was always bilateral. On slit-lamp examination, the anterior segment was normal, with no detectable signs of interface inflammation. Patients were given dexamethasone eyedrops (Maxidex; Alcon Laboratories, Inc.) six times daily for 2 weeks. Photophobia improved a few days after treatment.
All 10 eyes with TLSS had a UCVA of at least 20/25 at the final follow-up. There was no statistically significant difference between the preoperative and postoperative mean BSCVA (P=.343, paired t-test). One eye lost one line of BSCVA by the final follow-up (from 1.0?0.9). Figure 1 shows the visual acuity and refractive outcomes in eyes with TLSS.
In three patients having TLSS, we diagnosed unilateral grade 2 diffuse lamellar keratitis (DLK, 33.3%) in the first week after surgery. The incidence of DLK in eyes that did not develop TLSS was 3% (23 eyes). There was a statistically significant difference in DLK incidence between eyes with and without TLSS (P<.001). In all cases, DLK was unilateral and responded well to topical TobraDex every 2 hours for 7 days.
In the first 287 eyes operated on with the femtosecond laser, we used a low regimen of topical steroids. Eight (2.8%) developed TLSS. In the remaining 478 eyes, we used a high-dose regimen of topical steroids, and two (0.4%) developed TLSS. This represents a sevenfold decrease in the incidence of TLSS after postoperative topical steroids were increased (P<.001). Regarding other variables (ie, color of the iris, scotopic and mesopic pupil diameter, presence of opacified bubble layer), there were no significant differences when comparing healthy eyes with TLSS cases. Confocal microscopy was performed in seven of the 10 cases suffering TLSS; we detected a greater keratocyte activation degree in the cornea than after normal LASIK (Figure 2).
DISCUSSION
From our results, an inflammatory origin of TLSS is suggested. Compared with mechanical microkeratomes, the femtosecond laser had been associated with a higher postoperative inflammatory reaction and fibrosis adjacent to the flap margin.19 Stonecipher et al17 observed a fivefold reduction of TLSS when laser energy settings were lowered by 20%. The investigators also found that lower energy settings during femtosecond laser ablation may prevent keratocyte activation.
In our series, a significantly decreased incidence of TLSS occurred after we increased the postoperative steroid regimen; we continued to use similar levels of energy in flap creation. DLK was 10 times more frequent in eyes that developed TLSS (30%) versus eyes that did not (3%), suggesting that increased inflammation postoperatively may increase the incidence of TLSS. Grade 2 DLK developed unilaterally in three patients. Endotoxin-related DLK is almost never unilateral, although it is often highly asymmetric.20 Therefore, DLK in eyes that developed TLSS may not have been the typical endotoxin DLK, but rather another form of inflammation associated with femtosecond energy. Even in cases of unilateral DLK, patients reported light sensitivity in both eyes, not only in the eye with previous DLK.
In summary, the combined use of the lowest possible energy and an intensive postoperative topical steroid regimen for the first few days after femtosecond laser flap creation helped decrease the incidence of TLSS. Further clinical studies may identify more strategies for the prevention and management of this complication.
ENERGY-RELATED DLK
We also used the confocal microscope to compare the quality of flaps made with a third-generation mechanical microkeratome (Moria M2; Moria, Antony, France) to those obtained with a femtosecond laser (IntraLase).21
The prospective, consecutive, masked investigation was performed at the same surgical facility between October and December 2004. Results are from 200 eyes (100 patients) that underwent LASIK to correct myopia or myopic astigmatism. Eyes were assigned to the femtosecond laser or microkeratome group. The first 20 consecutive eyes (10 patients) of each group were studied using confocal microscopy.
The Moria microkeratome used a cutting head with a -1 ring, taking into account the keratometry values with the stop for creating an 8-mm flap. (The Moria microkeratome has different suction ring sizes to be used depending on the curvature of the cornea.) The 130-µm head created a desired flap thickness of 160 µm, with a 4.5-mm superior hinge. After checking adequate suction using an applanation tonometer, one blade was used for each eye. The femtosecond laser eyes were operated on using the 15-kHz engine, an attempted depth of 120 µm, attempted hinge length of 50º corresponding to a hinge length of 3.8 mm, and an attempted 9-mm flap.
After bilateral surgery was performed, surface measurements (ie, horizontal and vertical flap diameters, hinge length of the created flap) were registered using a surgical caliper under the surgical microscope view.
The laser ablation and postoperative care were similar in all cases. LASIK ablations were performed with the Esiris excimer laser (Schwind eye-tech-solutions, Kleinostheim, Germany), and the interface was irrigated with 3 mL of balanced salt solution (BSS; Alcon Laboratories, Inc.) using a bimanual and automated I/A system. Postoperative treatment consisted of Tobradex every 6 hours during week 1 and preservative-free sodium hyaluronate tears (Vislube; Thea Laboratory, Barcelona, Spain) every 12 hours for 1 month.
At postoperative months 1 and 3, flap thickness, particle density, and the wound healing opacity (WHO) index were evaluated using a tandem scanning confocal microscope (TSCM Model 165A; Tandem Scanning, Reston, Virginia). Clarity of an operated cornea is related to the activation of keratocytes and the creation of collagen and extracellular matrix by fibroblast. WHO index evaluates the healing process by quantifying the amount of reflected light at the anterior part of the corneal stroma.22-24
RESULTS
Eyes operated on with either the femtosecond or mechanical microkeratome did not present significant differences regarding visual and refractive outcomes, and the safety, efficacy, and predictability indices for both groups were also similar. Respecting intraoperative flap recordings, the difference between the attempted and achieved flap diameters was statistically significant for the vertical meridian (t-test; P=.016). The femtosecond laser performed more round and less oval flaps compared with the mechanical microkeratome.
At week 1, grades one and two DLK were observed in 16 eyes of the IntraLase group. In the same group, grade three DLK developed with an impact on its refractive outcome. No eye operated on using the mechanical microkeratome developed this condition (P<.001).
Regarding the safety for the DLK subgroup, only the patient with grade three DLK lost more than two lines of BSCVA at 3 months postoperative, and no significant differences in BSCVA were found between the femtosecond and mechanical microkeratome groups (P=.54).
After analyzing the study results, we decided to reduce the energy for the raster cut from 1.6 to 1.2 µJ. Our DLK rate then decreased from 16% to 4%. Figure 3 shows the granulated infiltrative pattern that the grade three DLK case presented at different levels of the interface and anterior corneal stroma, as examined by confocal microscopy. There were no other intra- or postoperative complications.
The thickness of the flap, the absolute value of the differences between the desired and actual performed flap thickness, the density of the particles at the central part of interfaces, and the degree of opacity related to the WHO index estimated for both groups at the first and third month after surgical procedures are shown in Table 1.
To study the quality of the flap created by different models of microkeratomes, confocal microscopy has been proved useful,21-23 and the TSCM precisely determines the total corneal thickness and the different corneal sublayers.25-27 It is also useful for measuring corneal flap thickness after LASIK.28
Backscattered light emanating from the anterior corneal stroma indicates a lack of transparency, which is associated with an activation of the anterior keratocytes and production of extracellular matrix and collagen following LASIK.29 This activation is known to be a normal process in the healing response of the corneal stroma. It is greater as the inflammation induced by the wound increases. We termed the parameter that quantifies this backscattered light as the confocal wound healing opacity index, in which a high index would indicate an evident or even subclinical lack of transparency of the operated cornea. This may be considered as an indirect indicator of the inflammation.
Decreased contrast sensitivity has been found in eyes with a high WHO index after LASIK.30 We used the confocal WHO index and found a trend toward an elevated index in the eyes in the femtosecond laser group, which may be explained by the deeper cuts performed with the M2 microkeratome and a greater degree of inflammation observed in the femtosecond laser group.
We found a lower density of particles at the interfaces of eyes operated on by IntraLase?as we expected. This is perhaps due to the absence of a metallic blade and a smaller degree of manipulation of the anterior corneal stroma. The progressive disappearance of the particles could be explained by their predominantly organic nature.
DLK consists of a multietiologic syndrome that expresses variable degrees of inflammation at the interface of eyes that underwent LASIK.20,30 As there were no reports of DLK with the femtosecond laser, the prevalence of this syndrome in our series has to be taken into account.
The cases of DLK almost disappeared in our refractive surgery unit when we reduced the energy of the femtosecond laser. We now believe that the incidence of DLK after using IntraLase may be related to the intensity of the energy used for the lamellar or sidecut. From our point of view, each unit of this femtosecond cutting system should be monitored during installation to determine the optimal and safest level of energy delivered to the cornea.
The high prevalence of this incidence did not, however, have an impact on the refractive safety of the technique. The median BSCVA at 3 months postoperative for eyes that suffered any degree of DLK was not significantly different from the eyes that did not present with this syndrome. The complete absence of any other intra- or postoperative complications in the eyes operated on using IntraLase indicates a good level of safety for this device.
Two new inflammatory conditions, TLSS and energy-related DLK, appeared specifically related with the use of the femtosecond laser. From the results of our two studies, we concluded that although the femtosecond laser was a safe device for creating lamellar cuts for LASIK, the use of higher energy levels could induce a greater interface inflammatory reaction. Two strategies proved to be effective to reduce, at least in part, the incidence of TLSS and DLK: (1) adjusting the parameters of the femtosecond laser to reduce the amount of energy used to make the lamellar resection and (2) increasing the dose of steroids in the early postoperative period.
César Albarrán-Diego practices in the Refractive Surgery Department, Centro de Especialidades Marqués de Sotelo, and Hospital NISA Virgen del Consuelo, in Valencia, Spain. Mr. Albarrán-Diego states that he has no financial interest in the products or companies mentioned. He may be reached at cesar.albarran@gmail.com.
Jorge L. Alió, MD, PhD, is Professor and Chairman of Ophthalmology, Miguel Hernández University, Alicante, Spain, and Medical Director of VISSUM Corp., in Spain. Professor Alió states that he has no financial interest in the products or companies mentioned. He may be reached at +34 96 515 00 25; jlalio@vissum.com.
Jaime Javaloy, MD, PhD, practices in the Department of Refractive Surgery, VISSUM Instituto Oftalmológico de Alicante, and Division of Ophthalmology, Miguel Hernández University, Medical School, in Alicante, Spain. Dr. Javaloy states that he has no financial interest in the products or companies mentioned. He may be reached at tel: +34 96 515 00 25; fax: +34 96 515 15 01; or jjavaloy@coma.es.
Gonzalo Muñoz, MD, PhD, practices in the Department of Surgery, VISSUM Instituto Oftalmológico de Alicante, Division of Ophthalmology, Miguel Hernández University, Medical School, in Alicante, Spain, and Refractive Surgery Department, Centro de Especialidades Marque's de Sotelo and Hospital NISA Virgen del Consuelo, in Valencia, Spain. Dr. Muñoz states that he has no financial interest in the products or companies mentioned. He may be reached at gon.munoz@ono.com.
María T. Vidal, MD, PhD, practices in the Department of Refractive Surgery, VISSUM Instituto Oftalmológico de Alicante, Division of Ophthalmology, Miguel Hernández University, Medical School, in Alicante, Spain, and Department of Ophthalmology, Hospital General Universitario de Elche, in Spain. Dr. Vidal states that she has no financial interest in the products or companies mentioned. She may be reached at jjavaloy@coma.es.
