ARTICLE

Mechanical epithelial removal followed by corneal collagen crosslinking in progressive keratoconus: Short-term complications Nurullah Cagil, MD, Ozge Sarac, MD, Hasan Basri Cakmak, MD, Gamze Can, MD, Erol Can, MD

PURPOSE: To compare the complications occurring within the first 3 months of corneal collagen crosslinking (CXL) performed with mechanical or transepithelial phototherapeutic keratectomy (PTK) epithelial removal in keratoconus patients. SETTING: Yildirim Beyazit University Ataturk Training and Research Hospital, Ankara, Turkey. DESIGN: Nonrandomized retrospective clinical study. METHODS: Eyes of consecutive progressive keratoconus patients who had PTK or mechanical epithelial removal followed by CXL were included. All patients were examined regularly until epithelial healing. Detailed ophthalmologic examinations were performed preoperatively and 1 and 3 months postoperatively. RESULTS: The study comprised 499 eyes (302 patients) that had transepithelial PTK (Group 1, 153 eyes) or mechanical epithelial removal (Group 2, 256 eyes) followed by CXL. Delayed epithelial healing occurred in 15.0% of eyes in Group 1 and 3.5% of eyes in Group 2 (P Z .001). Epithelial hypertrophy occurred in 24.8% of eyes and 3.5% of eyes, respectively (P Z .001). Salzmann-like epithelial nodules (2.6%), epithelial herpetic keratitis (1.9%), anterior uveitis (1.9%), and elevated intraocular pressure (1.9%) occurred in Group 1 only and infective keratitis (0.8%) in Group 2 only. Marked stromal edema and peripheral sterile infiltrates occurred at similar rates in both groups (P Z .567 and P Z .479, respectively). Grade 1C corneal haze was significantly high in Group 2. Grade 2C and 3C haze was significantly high in Group 1 (P Z .001). CONCLUSIONS: Ocular surface healing disorders were the most common early complications of CXL. Short-term complications were higher with the transepithelial PTK epithelial removal technique than with mechanical epithelial removal. Financial Disclosure: No author has a financial or proprietary interest in any material or method mentioned. J Cataract Refract Surg 2015; 41:1730–1737 Q 2015 ASCRS and ESCRS

Numerous therapeutic approaches have been used in the management of keratoconus including spectacles, contact lenses, intrastromal corneal ring segment implantation, and lamellar or penetrating corneal transplantation surgery.1–3 Since 2003, these treatment strategies have been supplemented by corneal collagen crosslinking (CXL), a procedure performed to stop or decrease the progression of keratoconus by increasing the biomechanical strength and stability of the cornea.4 Unlike the previously mentioned 1730

Q 2015 ASCRS and ESCRS Published by Elsevier Inc.

therapeutic options, CXL is a unique approach designed to arrest the progression of the corneal ectasia. Corneal CXL enhances corneal rigidity by increasing the degree of covalent bonding between and within the molecules of collagen fibrils and the proteoglycans using riboflavin and ultraviolet-A (UVA) light.5,6 Today, the accepted protocol for CXL includes deepithelialization of the cornea before the administration of riboflavin to increase its penetration throughout the corneal stroma and thus achieve a high http://dx.doi.org/10.1016/j.jcrs.2014.12.058 0886-3350

SHORT-TERM COMPLICATIONS OF CORNEAL CXL

level of UVA absorption.7 Traditionally, the deepithelialization procedure has been performed mechanically. In a new excimer laser transepithelial phototherapeutic keratectomy (PTK) protocol, laser ablation is used to remove the epithelium and anterior irregular corneal stroma before CXL is performed.8,9 It has been suggested that epithelial removal using transepithelial PTK before CXL results in better visual and refractive outcomes than mechanical epithelial removal in keratoconus patients.8,9 Ultraviolet-A irradiation has a toxic effect on cell viability and can cause keratocyte and corneal endothelial cell destruction or death as well as possible lens and retinal damage during the CXL procedure.10,11 Ocular surface-healing disorders or secondary ocular surface infections can also occur after CXL.12,13 Decreased vision, delayed epithelial healing, hypertrophic epithelial healing, corneal haze, endothelial damage, marked corneal edema, microbial keratitis, peripheral sterile infiltrates, herpes reactivation, corneal scarring, intraocular pressure (IOP) elevation, and treatment failure are documented complications of CXL.12–17 In the present study, we evaluated the early complications (occurring within the first 3 months) of CXL performed with mechanical epithelial removal or transepithelial PTK epithelial removal in patients with progressive keratoconus. The effects of the epithelial debridement technique during CXL on the rates of the complications were assessed. PATIENTS AND METHODS This single-center nonrandomized retrospective study was performed in compliance with the institutional and government review board regulations, informed consent regulations, and the tenets of the Declaration of Helsinki. All patients provided written informed consent. The study included the eyes of consecutive progressive keratoconus patients who were between the ages of 10 years and 45 years who had CXL performed after transepithelial PTK epithelial removal (Group 1) or after mechanical epithelial removal (Group 2). Patients who had a central corneal thickness (CCT) at the thinnest point thicker than 450 mm associated with a manifest refraction spherical equivalent

Submitted: July 16, 2014. Final revision submitted: December 4, 2014. Accepted: December 28, 2014. From the Department of Ophthalmology, Yildirim Beyazit University, Ankara Ataturk Training and Research Hospital, Ankara, Turkey. Corresponding author: Ozge Sarac, MD, Yildirim Beyazit University, Ankara Ataturk Training and Research Hospital, Department of Ophthalmology, Bilkent, Ankara, Turkey. E-mail: drozgesarac@ gmail.com.

1731

(SE) greater than 3.00 diopters (D), an uncorrected distance visual acuity (UDVA) worse than 0.8 decimal Snellen fraction (0.1 logMAR equivalent), and rigid gas-permeable contact lens intolerance were chosen to form Group 1. Patients who had 1 of the following criteria were placed in Group 2: a CCT at the thinnest point thinner than 450 mm, SE less than 3.00 D, or UDVA better than 0.8 decimal. Keratoconic eyes were diagnosed clinically if they had at least 1 clinical sign other than the topographic appearance of the map, which included Munson sign, scissors reflex during retinoscopy, corneal thinning, Fleischer ring, Vogt striae, increased visibility of the corneal nerves, and Rizzutti sign. The criteria for the progression of the keratoconus were an increase in the maximum keratometry (K) value of 1.00 D in 1 year measured with a Sirius 3-dimensional (3-D) rotating Scheimpflug camera and topography system (Costruzione Strumenti Oftalmici), subjective reports of deteriorating corrected distance visual acuity (CDVA), or the need for new contact lens fitting more than once in 2 years. Exclusion criteria were a corneal thickness less than 400 mm, the presence of a central or paracentral corneal scar, a history of herpetic keratitis, active ophthalmic inflammation or infection, pregnancy or lactation, and severe dry eye before CXL. In the early postoperative period, all patients were examined at 1 to 2 days until epithelial healing. If the borders of the corneal epithelial wound were not determined under the bandage contact lens during slitlamp examination, the contact lens was removed and an examination was performed after instillation of fluorescein 2.0%. The ophthalmologic examination consisted of visual acuity testing including UDVA and CDVA, slitlamp evaluation, IOP measurement with applanation tonometry, fundoscopic evaluation, and topographic analysis of the cornea. These examinations were performed before CXL and 1 month and 3 months after CXL. Patients who had any CXL complication were examined more frequently. Topographic analysis was performed with the rotating Scheimpflug camera and topography system, which is a diagnostic device that combines a 3-D rotating Scheimpflug camera with a Placido disk topographer.18 From the topographic analysis, the simulated K1, simulated K2, average K, CCT, and CCT at the thinnest point were evaluated. Significant visual loss was defined as a loss of 3 or more Snellen lines of CDVA at the 3-month follow-up compared with baseline. Three months after CXL, corneal haze was evaluated and graded with slitlamp biomicroscopy on a scale from 1 to 4.19 In the present study, corneal haze was not considered a short-term complication. It was graded as follows: 0C Z clear cornea; 1C Z focal areas of corneal clouding or reticulation; 2C Z diffuse mild stromal clouding or reticulation; 3C Z diffuse stromal clouding or reticulation obscuring view of iris details; 4C Z focal or diffuse areas of dense stromal haze obscuring iris detail.

Surgical Technique All procedures were performed in an operating room under sterile conditions. After topical anesthesia of proxymetacaine hydrochloride 0.5% eyedrops (Alcaine) was administered, the corneal epithelium was removed. In Group 1, the corneal epithelium was removed by transepithelial PTK. The Esiris scanning-spot laser (Schwind eye-tech-solutions GmbH & Co.) was used for the procedure.

J CATARACT REFRACT SURG - VOL 41, AUGUST 2015

1732

SHORT-TERM COMPLICATIONS OF CORNEAL CXL

A single-step transepithelial PTK ablation was performed in an 8.5 mm zone at a constant depth of 50 mm. Laser profiles were not based on the refractive parameters and were not topography- or wavefront-guided. Mitomycin-C was not used. In Group 2, the corneal epithelium was removed mechanically using a crescent knife at an intended 8.5 mm zone after the corneal epithelium was loosened with an alcohol 20% solution in an 8.5 mm alcohol well. The alcohol was applied for 20 seconds over the cornea. After epithelial removal, the residual corneal thickness was measured with an ultrasonic pachymeter (Palmscan AP-2000-Ultima, Micromedical Devices, Inc.). Riboflavin drops were applied on the center of the cornea every 3 minutes for 30 minutes until the aqueous stained yellow. If the CCT at the thinnest point was measured below 400 mm after epithelial removal, hypotonic riboflavin (riboflavin 0.1% in sodium chloride 0.009%, Meran Tıp Ltd.) was used. If the CCT at the thinnest point was above 400 mm, isotonic riboflavin (riboflavin 0.1% in dextran 20.0% T500 solution, Meran Tıp Ltd.) was used. Ultraviolet-A irradiation was accomplished using a commercially available UVA system (Meran Tıp Ltd.). Before treatment, the intended 3 mW/cm2 surface irradiance (5.4 J/cm2 surface dosage after 30 minutes) was calibrated using a UVA meter (UVA-365, Lutron Electronic Enterprise Co. Ltd.). During treatment, riboflavin solution was applied every 2 minutes to ensure saturation and a balanced salt solution was applied every minute to moisten the cornea. Ultrasonic pachymetry measurements were taken 2 times (15 minutes and 30 minutes) after the start of riboflavin drops. Topical anesthetic drops were instilled every 10 minutes throughout the procedure. A silicone hydrogel bandage contact lens (Acuvue Oasis, Johnson & Johnson, Vision Care) was applied at the end of the surgery until full reepithelialization of the cornea. Postoperative treatment included ofloxacin eyedrops (Exocin) 4 times a day for 1 week, fluorometholone eyedrops (FML) 4 times a day on a tapering schedule for 1 month, and artificial tears 4 times a day for 6 months. Delayed epithelial healing was diagnosed when healing had not occurred more than 5 days after CXL.

Statistical Analysis Data analysis for each outcome parameter was performed using SPSS software (version 18.0, International Business Machines Corp.). The statistical software G* power (version 3.2.9.2) was used for post hoc power analysis.20 All values are expressed as the mean G SD and range. The chi-square test was used to assess differences in categorical variables. The independent t test was used to assess differences in scale variables. A P value less than 0.05 was considered statistically significant. Visual acuity was expressed in logMAR notation for the statistical analysis.

RESULTS Four hundred nine eyes of 302 keratoconus patients were included in this study. There were 129 (42.7%) women and 173 (57.3%) men with a mean age of 23.57 G 6.2 years (range 10 to 43 years). Group 1 consisted of 153 eyes of 114 patients. There were 51 women (44.7%) and 63 men (52.9 %) in Group 1. Group 2 included 256 eyes of 188 patients of whom 78 (41.5%) were women and 110 (58.5%) were men. Table 1 shows

the mean baseline topographic data of groups including simulated K1, simulated K2, average K, CCT, CCT at the thinnest point, and the mean SE and cylindrical refractive values. All patients had uneventful CXL treatment. Short-Term Complications Table 2 shows the short-term complications in Group 1 and Group 2. The most frequent complications were corneal epithelial hypertrophy (defined as elevated and disorganized epithelium located in the central or paracentral cornea) and delayed epithelial healing (Figure 1). Statistically significantly more eyes in Group 1 than in Group 2 had these 2 complications (P Z .001, 1-b Z 0.99 and P Z .001, 1-b Z 0.97, respectively). The mean epithelium healing time was 5.31 G 1.72 days in Group 1 and 5.09 G 0.77 days in Group 2 (P Z .08, 1-b Z 0.36). In Group 1, epithelial healing occurred in 130 eyes (85.0%) by 5 days, in 1 eye (0.7%) at 6 days, in 2 eyes (1.3%) at 7 days, in 8 eyes (5.2%) at 8 days, and in 4 eyes (2.6%) at 10 days; in 8 eyes (5.2%), it took more than 10 days. In Group 2, epithelial healing occurred in 247 eyes (96.5%) by 5 days, in 2 eyes (0.8%) at 7 days, and in 5 eyes (1.9%) at 10 days; in 2 eyes (0.8%), it took more than 10 days. Salzmann-like epithelial nodules were present in Group 1 only (Figure 2) and persisted for a mean of 1 month. Corneal peripheral sterile infiltrates occurred in an equal number of eyes in Group 1 and in Group 2 (P Z .479) (Figure 3). They resolved within 1 month in all cases. In Group 1, no eye had infectious keratitis. In Group 2, 2 eyes (0.3%) had infectious keratitis; 1 was culture positive (Pseudomonas aeruginosa). Marked stromal edema occurred in both groups (P Z .567). It persisted for a mean duration of 1 month in both groups. Anterior uveitis occurred in Group 1 Table 1. Baseline topographic data of eyes in Group 1 and Group 2. Mean G SD Parameter

Group 1 (n Z 153)

Group 2 (n Z 256)

Simulated K1 (D) 48.62 G 4.91 52.99 G 4.38 Simulated K2 (D) 52.22 G 5.72 57.60 G 5.54 Mean K (D) 50.34 G 5.22 55.16 G 4.79 CCT (mm) 465.13 G 43.22 441.59 G 51.44 t-CCT (mm) 451.69 G 44.33 427.73 G 47.35 SE (D) 3.18 G 2.67 2.03 G 2.45 Cylinder (D) 3.79 G 1.08 4.87 G 1.15

P Value .001 .001 .001 .001 .001 .001 .001

CCT Z central corneal thickness; K Z keratometry; K1 Z keratometry in flat meridian; K2 Z keratometry in steep meridian; SE Z spherical equivalent; t-CCT Z central corneal thickness at the thinnest point

J CATARACT REFRACT SURG - VOL 41, AUGUST 2015

SHORT-TERM COMPLICATIONS OF CORNEAL CXL

1733

Table 2. Comparison of short-term complications in Group 1 and Group 2. Number (%) Complication

Group 1

Group 2

P Value

Delayed epithelial healing Epithelial hypertrophy Peripheral sterile infiltrate Microbial infiltrate Marked stromal edema Anterior uveitis Herpes reactivation Elevated intraocular pressure Salzmann-like epithelial nodule

23 (15.0) 38 (24.8) 4 (2.6) 0 5 (3.2) 3 (1.9) 2 (1.3) 3 (1.9)

9 (3.5) 9 (3.5) 4 (1.5) 2 (0.3) 9 (3.5) 0 1 (0.3) 0

.001 .001 .479 .391 .567 .052 .559 .052

0

.04

4 (2.6)

Figure 2. Salzmann-like epithelial nodule occurred after transepithelial PTK before corneal CXL.

only; all cases were detected at the 1-week follow-up examination. None of the patients had an ocular or systemic disorder that is associated with the etiology of uveitis. Herpetic keratitis occurred in Group 1 only (Figure 4). In all cases, the herpetic keratitis diagnosis was confirmed by a polymerase chain reaction analysis of the corneal swab for herpes simplex virus (HSV) a mean of 7 days postoperatively. Elevated IOP was detected in Group 1 only. All cases were thought to be topical steroid induced.

eye (0.4%). No eye developed 4C corneal haze. Grade 1C haze was significantly high in Group 2, while grade 2C and 3C haze was significantly high in Group 1 (P Z .001).

Corneal Haze In Group 1, corneal haze 3 months after CXL was graded as 0C in 1 eye (0.7%), 1C in 94 eyes (68.6%), 2C in 38 eyes (27.7%), and 3C in 4 eyes (2.9%). In Group 2, it was graded as 0C in 2 eyes (0.8%), 1C in 225 eyes (87.9%), 2C in 28 eyes (11%), and 3C in 1

DISCUSSION The present study documented and compared the short-term complications of CXL in eyes with progressive keratoconus. Two epithelial removal techniques were used; that is, transepithelial PTK or mechanical epithelial removal. Short-term complications occurred

Figure 1. Delayed epithelial healing occurred after corneal CXL performed with mechanical epithelium removal.

Figure 3. Peripheral sterile infiltrate in a patient who had mechanical epithelium removal before corneal CXL.

Visual Acuity None of the complications in the study resulted in a significant loss of UDVA or CDVA. Table 3 shows the mean UDVA and CDVA in both groups at baseline and 1 month and 3 months after CXL.

J CATARACT REFRACT SURG - VOL 41, AUGUST 2015

1734

SHORT-TERM COMPLICATIONS OF CORNEAL CXL

Figure 4. Herpetic epithelial keratitis that occurred after transepithelial PTK before corneal CXL.

in 37.3% of eyes that had transepithelial PTK and 9.4% of eyes that had mechanical epithelial removal 3 months after CXL treatment (more than 1 type of complication per eye). The most common complications in both groups were ocular surface healing disorders such as delayed epithelial healing and abnormal hypertrophic epithelial healing. Corneal CXL increases the biomechanical strength of the corneal tissue by exposing the ectatic cornea to riboflavin and UVA irradiation.4 Epithelial removal is an essential first step of CXL to achieve adequate and homogenous absorption of riboflavin into the corneal stroma.7 Over the past several years, corneal epithelial removal has been performed mechanically using a rotating brush or a blade in a standard CXL procedure.4 Recently, in a new transepithelial PTK protocol, epithelial removal has been performed using an excimer laser before CXL.8,9

Table 3. Mean UDVA and CDVA in Group 1 and Group 2 at baseline and 1 and 3 months after CXL. Visual Acuity (LogMAR) UDVA Baseline 1 month 3 months CDVA Baseline 1 month 3 months

Mean G SD Group 1

Group 2

P Value

0.27 G 0.02 0.25 G 0.02 0.35 G 0.02

0.21 G 0.01 0.20 G 0.01 0.25 G 0.01

.001 .001 .001

0.65 G 0.04 0.52 G 0.02 0.63 G 0.02

0.53 G 0.03 0.45 G 0.01 0.55 G 0.01

.001 .001 .001

CDVA Z corrected distance visual acuity; UDVA Z uncorrected distance visual acuity

Excimer laser epithelial removal is a well-known technique and has been successfully used before refractive surgery as well as to treat anterior corneal disorders such as corneal scars; corneal epithelial, Bowman membrane, and anterior stromal dystrophies; and elevated corneal lesions.21 The excimer laser was first used in conjunction with CXL treatment in keratoconus patients in 2007 in the Athens protocol (combined topography-guided photorefractive keratectomy [PRK] and CXL).22 This protocol aims to stop the progression of ectasia and decrease irregular astigmatism. Clinical studies22,23 report good safety records and effective refractive, keratometric, and topographic results with the Athens protocol. After the Athens protocol, a new transepithelial PTK protocol was introduced into the clinical practice in which a constant depth of 50 mm excimer laser ablation in the central zone of the cornea is performed.8,9 In contrast to the Athens protocol, the transepithelial PTK protocol has not been performed as a topography-guided procedure. In keratoconus, the corneal epithelium is known to be thin in the area overlying the cone that is surrounded by a region of epithelial thickness called the epithelial doughnut pattern.24 Because of this pattern, transepithelial PTK epithelial removal at a constant depth not only removes the epithelium but also ablates some stromal tissue at the apex of the cone besides the epithelium. This smoothing effect can lead to better visual and refractive outcomes in patients treated with transepithelial PTK–associated CXL.8,9 In PRK before CXL, depending on the epithelial thickness and the amount of myopia being treated, in general more corneal stromal tissue over the cone area is removed than with transepithelial PTK. In transepithelial PTK–associated CXL, an intended ablation depth of 50 mm, including corneal epithelium, might lead to less disruption of the biomechanical integrity of the cornea than PRK-associated CXL. Although CXL is considered a safe and welltolerated low-invasive procedure, complications in CXL treatment are reported to occur in 1% to 10% of patients.12 They can be primary as a result of the toxic effects caused by the procedure itself, incorrect application of the technique, or incorrect patient selection. Secondary complications can also occur related to application of the therapeutic soft contact lens after the procedure, poor patient hygiene, or associated ocular surface disease.13 Recent evidence suggests that compromised ocular surface healing after CXL is the most common complication of the procedure that can affect the final visual acuity.A,B In the present study, we evaluated the short-term complications of CXL performed with 2 epithelium removal techniques (transepithelial PTK and mechanical) in eyes with progressive keratoconus. The rates

J CATARACT REFRACT SURG - VOL 41, AUGUST 2015

SHORT-TERM COMPLICATIONS OF CORNEAL CXL

of the complications with both techniques were compared. To our knowledge, the current study is the first to compare the short-term complications of CXL with those of transepithelial PTK. Similar to previous reports,A,B in our study corneal epithelial healing disorders, such as delayed epithelial healing and epithelial hypertrophy, were the most common complications of both techniques. Some studies report that the corneal epithelium attains a regular morphology and density within 5 days after CXL.25 In a recent study,A delayed epithelial healing was reported in 4 of 23 eyes that had standard CXL. In our study, the rate of delayed epithelial healing was similar (3.5%) in the mechanical epithelial removal group. It was nearly 4 times higher in eyes that had transepithelial PTK before CXL, occurring in 15.0% of eyes. Epithelial hypertrophy occurred 7 times more often in eyes that had transepithelial PTK than in eyes that had mechanical epithelial removal. It was detected in 24.8% of eyes in the transepithelial PTK group and in 3.5% of eyes in the mechanical epithelial removal group. A previous studyA reports findings similar to ours; in that study, it occurred in 4 of 23 eyes that had mechanical epithelial removal before CXL. Salzmann-like epithelial nodules were seen in the transepithelial PTK group only and were present in 2.6% of eyes. Salzmann-like epithelial nodules, including 1 or multiple white epithelial nodular lesions, were first reported in eyes that had CXL associated with PRK in the Athens protocol.B Nodules were reported in 25% of patients who were treated using the Athens protocol.B This high percentage might be due to the deeper stromal ablation required for topography-guided PRK performed before CXL treatment in the Athens protocol. Keratoconic eyes with high K values before CXL have higher complication rates.26 In our study, keratoconic eyes in the transepithelial PTK group had higher complication rates than eyes in the mechanical epithelial removal group. Eyes before CXL in the transepithelial PTK group had flatter corneas than eyes treated in the mechanical epithelial removal group. During the transepithelial PTK procedure, a 50 mm excimer laser ablation was used to remove the epithelium and the apex of the cone, where the epithelium is thinner in keratoconic corneas than in normal corneas. The ablation of Bowman membrane and the stromal tissue over the apex of the cone and the exposure of the stroma to riboflavin and UVA might cause the ocular surface healing disorders, herpetic keratitis reactivation, and intraocular inflammation. Supporting the argument above, anterior uveitis, IOP elevation, and epithelial herpetic keratitis occurred in 1.9%, 1.9%, and 1.9% of eyes, respectively,

1735

in which transepithelial PTK was performed before CXL. No eye in the mechanical epithelial removal group had anterior uveitis, epithelial herpetic keratitis, or IOP elevation. Herpetic keratitis reactivation after standard CXL has been reported.16,27 It seems as though UVA light and the postoperative use of topical steroids could be potent stimuli that reactivate latent HSV infections, even in eyes without a history of clinical herpes virus ocular infections. Stromal trauma and excimer laser irradiation might be the additional risk factors for HSV reactivation in eyes that have transepithelial PTK before CXL. Our results indicate that excimer laser treatment did not adversely affect the development of sterile or infectious keratitis. Peripheral sterile corneal infiltrates were seen in similar rates in both groups (2.6% in Group 1; 1.5% in Group 2). Another study reports26 a 7.6% incidence in eyes that had CXL treatment with mechanical epithelial removal. Similar to results in previous reports,14,28 the rate of infectious keratitis was low in our patients. Only 2 eyes in the mechanical epithelial removal group presented with infectious keratitis. One was culture positive, and the causative agent was P aeruginosa. Infectious keratitis can occur after CXL because of the presence of an epithelial defect, soft bandage contact lens application, and the use of topical steroids immediately after the procedure. As CXL damages bacteria and fungi, contact with the infectious agent generally occurs in the early postoperative period. A typical corneal haze characterized by dust-like changes in the 60% depth of the corneal stroma with a midstromal demarcation line has generally been observed after CXL.19,29 The exact pathophysiology of the development of the corneal haze after CXL is not known yet. Loss of keratocytes in the corneal stroma,30 proteoglycan–collagen interactions,31 and glycosaminoglycan hydration32 might be causes. Corneal haze emerges in a few days after CXL and in general resolves over time. It increases during the first month, and this increase plateaus between the first and third months after CXL. After the third month, the cornea usually begins to clear.19 In the present study, corneal haze developed in 99% of patients who were treated with transepithelial PTK or mechanical epithelium removal before CXL. With both procedures, the most severe corneal haze was grade 3. The haze was more intense in patients who had photoablation before CXL. Our study has strengths and limitations. It strengths are the large number of patients and the comparison of the new excimer laser epithelial removal technique (transepithelial PTK) and the conventional technique (mechanical epithelial removal). The first limitation of our study is the heterogeneity of the groups. Group

J CATARACT REFRACT SURG - VOL 41, AUGUST 2015

1736

SHORT-TERM COMPLICATIONS OF CORNEAL CXL

1 consisted of more myopic eyes with thicker corneas, while Group 2 consisted of eyes with higher K values and thinner corneas. This heterogeneity might have affected the rates of the complications in each group. The second limitation is we did not evaluate the corneal endothelial cell layer before or after the CXL procedure. In our study, we detected marked stromal edema in 3.2% of eyes in the transepithelial PTK group and in 3.5% of eyes in the mechanical epithelial removal group. In these eyes, we could not rule out the presence of endothelial abnormalities because specular microscopy was not performed in these eyes before or after CXL. In conclusion, in the present study excimer laser transepithelial PTK epithelial removal before CXL led to higher complication rates than mechanical epithelial removal. Most short-term complications with both epithelial removal techniques were minor and occurred and resolved during the first month of surgery. Therefore, patients who had CXL with either epithelial removal technique would require close follow-up by an experienced cornea specialist during the first month after surgery. WHAT WAS KNOWN  Ocular surface healing disorders, such as delayed epithelial healing and epithelial hypertrophy, are the most common complications after standard CXL. WHAT THIS PAPER ADDS  The occurrence and rates of early postoperative complications, such as ocular surface healing, were higher with transepithelial PTK epithelial removal than with mechanical epithelial removal.  Herpetic keratitis reactivation, intraocular inflammation, and IOP elevation were more common in the transepithelial PTK epithelial removal group.

REFERENCES 1. Barnett M, Mannis MJ. Contact lenses in the management of keratoconus. Cornea 2011; 30:1510–1516 2. Park J, Gritz DC. Evolution in the use of intrastromal corneal ring segments for corneal ectasia. Curr Opin Ophthalmol 2013; 24:296–301 3. Cassidy D, Beltz J, Jhanji V, Loughnan MS. Recent advances in corneal transplantation for keratoconus. Clin Exp Optom 2013; 96:165–172. Available at: http://onlinelibrary.wiley.com/doi/10. 1111/cxo.12047/pdf. Accessed June 15, 2015 4. Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-A– induced collagen crosslinking for the treatment of keratoconus. Am J Ophthalmol 2003; 135:620–627. Available at: http://grmc. ca/assets/files/collagen_crosslinking_2003_wollensak.pdf. Accessed June 15, 2015

5. Beshtawi IM, O’Donnell C, Radhakrishnan H. Biomechanical properties of corneal tissue after ultraviolet-A–riboflavin crosslinking. J Cataract Refract Surg 2013; 39:451–462 6. Jankov MR II, Jovanovic V, Nikolic L, Lake JC, Kymionis G, Cockunseven E. Corneal collagen cross-linking. Middle East Afr J Ophthalmol 2010; 17:21–27. Available at: http://www.ncbi. nlm.nih.gov/pmc/articles/PMC2880370/?reportZprintable. Accessed June 15, 2015 7. Baiocchi S, Mazzotta C, Cerretani D, Caporossi T, Caporossi A. Corneal crosslinking: riboflavin concentration in corneal stroma exposed with and without epithelium. J Cataract Refract Surg 2009; 35:893–899 8. Kapasi M, Baath J, Mintsioulis G, Jackson WB, Baig K. Phototherapeutic keratectomy versus mechanical epithelial removal followed by corneal collagen crosslinking for keratoconus. Can J Ophthalmol 2012; 47:344–347 9. Kymionis GD, Grentzelos MA, Kounis GA, Diakonis VF, Limnopoulou AN, Panagopoulou SI. Combined transepithelial phototherapeutic keratectomy and corneal collagen crosslinking for progressive keratoconus. Ophthalmology 2012; 119:1777–1784 10. Wollensak G, Spoerl E, Reber F, Seiler T. Keratocyte cytotoxicity of riboflavin/UVA-treatment in vitro. Eye 2004; 18:718– 722. Available at: http://www.nature.com/eye/journal/v18/n7/ pdf/6700751a.pdf. Accessed June 15, 2015 11. Wollensak G, Spoerl E, Wilsch M, Seiler T. Endothelial cell damage after riboflavin–ultraviolet-A treatment in the rabbit. J Cataract Refract Surg 2003; 29:1786–1790 12. Seiler TG, Schmidinger G, Fischinger I, Koller T, Seiler T. Complications of corneal cross-linking. Ophthalmologe 2013; 110:639–644 13. Dhawan S, Rao K, Natrajan S. Complications of corneal collagen cross-linking. J Ophthalmol 2011; Article ID:869015. Available at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3255294/pdf/ JOP2011-869015.pdf. Accessed June 15, 2015 14. Pollhammer M, Cursiefen C. Bacterial keratitis early after corneal crosslinking with riboflavin and ultraviolet-A. J Cataract Refract Surg 2009; 35:588–589 15. Raiskup F, Hoyer A, Spoerl E. Permanent corneal haze after riboflavin-UVA-induced cross-linking in keratoconus. J Cataract Refract Surg 2009; 25:S824–S828 16. Kymionis GD, Portaliou DM, Bouzoukis DI, Suh LH, Pallikaris AI, Markomanolakis M, Yoo SH. Herpetic keratitis with iritis after corneal crosslinking with riboflavin and ultraviolet A for keratoconus. J Cataract Refract Surg 2007; 33:1982–1984 17. Koppen C, Vryghem JC, Gobin L, Tassignon MJ. Keratitis and corneal scarring after UVA/riboflavin cross-linking for keratoconus. J Refract Surg 2009; 25:S819–S823 18. Savini G, Barboni P, Carbonelli M, Hoffer KJ. Repeatability of automatic measurements by a new Scheimpflug camera combined with Placido topography. J Cataract Refract Surg 2011; 37:1809–1816 19. Greenstein SA, Fry KL, Bhatt J, Hersh PS. Natural history of corneal haze after collagen crosslinking for keratoconus and corneal ectasia: Scheimpflug and biomicroscopic analysis. J Cataract Refract Surg 2010; 36:2105–2114. Available at: http://www.vision-institute.com/UserFiles/File/CXL%20Haze% 20Published.pdf. Accessed June 15, 2015 20. Faul F, Erdfelder E, Buchner A, Lang AG. Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods 2009; 41:1149–1160. Available at: http://www.gpower.hhu.de/fileadmin/redaktion/Fakultaeten/ Mathematisch-Naturwissenschaftliche_Fakultaet/Psychologie/ AAP/gpower/GPower31-BRM-Paper.pdf. Accessed June 15, 2015

J CATARACT REFRACT SURG - VOL 41, AUGUST 2015

1737

SHORT-TERM COMPLICATIONS OF CORNEAL CXL

21. Rapuano CJ. Excimer laser phototherapeutic keratectomy. Curr Opin Ophthalmol 2001; 12:288–293 22. Kanellopoulos AJ, Binder PS. Management of corneal ectasia after LASIK with combined, same-day, topography-guided partial transepithelial PRK and collagen cross-linking: the Athens protocol. J Refract Surg 2011; 27:323–331 23. Kanellopoulos AJ, Asimellis G. Keratoconus management: longterm stability of topography-guided normalization combined with high-fluence CXL stabilization (the Athens Protocol). J Refract Surg 2014; 30:88–93 24. Reinstein DZ, Archer TJ, Gobbe M. Corneal epithelial thickness profile in the diagnosis of keratoconus. J Refract Surg 2009; 25:604–610 25. Mazzotta C, Traversi C, Baiocchi S, Sergio P, Caporossi T, Caporossi A. Conservative treatment of keratoconus by riboflavin-UVA-induced cross-linking of corneal collagen: Qualitative investigation of corneal epithelium and subepithelial nerve plexus regeneration by in vivo HRT II system confocal microscopy in humans. Eur J Ophthalmol 2006; 16:530–535. Available at: http://medlib.yu.ac.kr/eur_j_oph/ejo_pdf/2006_ 16_530-535.pdf. Accessed June 15, 2015 26. Koller T, Mrochen M, Seiler T. Complication and failure rates after corneal crosslinking. J Cataract Refract Surg 2009; 35:1358–1362 27. Yuksel N, Bilgihan K, Hondur AM. Herpetic keratitis after corneal collagen cross-linking with riboflavin and ultraviolet-A for progressive keratoconus. Int Ophthalmol 2011; 31:513–515 28. Hafezi F. Significant visual increase following infectious keratitis after collagen cross-linking. J Refract Surg 2012; 28:587–588 29. Seiler T, Hafezi F. Corneal cross-linking-induced stromal demarcation line. Cornea 2006; 25:1057–1059 30. Wollensak G, Spoerl E, Wilsch M, Seiler T. Keratocyte apoptosis after corneal collagen cross-linking using riboflavin/UVA treatment. Cornea 2004; 23:43–49 31. Michelacci YM. Collagens and proteoglycans of the corneal extracellular matrix. Braz J Med Biol Res 2003; 36:1037–1046.

Available at: http://www.scielo.br/pdf/bjmbr/v36n8/4920.pdf. Accessed June 15, 2015 32. Wollensak G, Aurich H, Pham D-T, Wirbelauer C. Hydration behavior of porcine cornea crosslinked with riboflavin and ultraviolet A. J Cataract Refract Surg 2007; 33:516–521

OTHER CITED MATERIAL A. Wajnsztajn D, Frenkel S, Frucht-Pery J, “Early Complications After Crosslinking for Keratoconus,’’ poster presented at the joint meeting of the American Academy of Ophthalmology and Asia-Pacific Academy of Ophthalmology, Chicago, Illinois, USA, November 2012 B. Cho MY, Kanellopoulos K, “Short and Long-term Complications of Combined Topography-guided Photorefractive Keratectomy and Riboflavin/Ultraviolet A Corneal Collagen Cross-linking (the Athens Protocol) in 412 Keratoconus Eyes,” presented at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, USA, May 2011. ARVO E-Abstract 5202. Available at: http://www.abstracts online.com/plan/ViewAbstract.aspx?mIDZ2684&sKeyZ642d 1697-8b41-442c-80a6-2aa1d91d9d99&cKeyZ56fe9e1f-002f4bd1-87ba-b6e6a3316361&mKeyZ6f224a2d-af6a-4533-8bbb6a8d7b26edb3. Accessed June 15, 2015

J CATARACT REFRACT SURG - VOL 41, AUGUST 2015

First author: Nurullah Cagil, MD Department of Ophthalmology, Yildirim Beyazit University, Ankara Ataturk Training and Research Hospital, Ankara, Turkey