Corneal cross-linking (CXL) is a well-established surgical technique using riboflavin and ultraviolet-A (UVA) for the treatment of keratoconus to strengthen the corneal tissue, resulting in stabilization of the ectatic disorder.1–4 According to the Dresden protocol, the epithelium has to be removed during CXL to permit the penetration of riboflavin solution and the saturation of the corneal stroma.1 CXL without epithelial removal may impair the effectiveness of the procedure.5–7
Epithelial removal during CXL may be performed using several methods, such as mechanical debridement, use of alcohol, and excimer laser transepithelial phototherapeutic keratectomy (t-PTK). Combined t-PTK and CXL (Cretan protocol) seems to result in better visual and refractive outcomes than the standard procedure with mechanical epithelial debridement.8
The aim of the current study was to compare the long-term results of CXL for the treatment of keratoconus using two different techniques for epithelial removal: t-PTK versus mechanical epithelial debridement.
Patients and Methods
This prospective, comparative, interventional case series enrolled patients with progressive keratoconus. Inclusion criteria were progressive keratoconus, patient aged 18 years or older, and corneal thickness of greater than 400 µm. Keratoconus was described as progressive when there was an increase in the cone apex keratometry of 0.75 diopters (D) or alteration of 0.75 D in the spherical equivalent refraction in the past 6 months. Exclusion criteria were history of intraocular or corneal surgery, herpetic keratitis, and any other corneal or anterior segment pathological signs, pregnancy, and lactation.
All patients underwent CXL treatment. Fifteen eyes (13 patients; 10 men and 3 women) with a mean age of 26 ± 3.75 years (range: 20 to 32 years) underwent epithelial removal with t-PTK (Cretan protocol) and 15 eyes (13 patients; 9 men and 4 women) with a mean age of 26 ± 3.34 years (range: 21 to 32 years) underwent mechanical epithelial debridement (Dresden protocol) using a rotating brush during CXL. All eyes met the inclusion criteria. The epithelial removal technique was performed with the first eye receiving mechanical epithelial debridement and then the two techniques were used alternately regarding the eyes undergoing CXL treatment.
Institutional review board committee approval was obtained and all patients were appropriately informed before their participation in the study about the possible outcomes and the current clinical experience. Participants provided written informed consent in accordance with institutional guidelines, according to the Declaration of Helsinki.
Preoperative and postoperative (at 1, 2, 3, and 4 years) evaluation consisted of ocular and general health history assessment; autorefractometry (Canon autorefractor; Canon USA Inc., Lake Success, NY); corneal topography (iTrace, Tracey Technologies, Houston, TX; Galilei, Ziemer Ophthalmic Systems AG, Biel, Switzerland); uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), manifest refraction, endothelial cell density (ECD) obtained with Tomey EM-3000 specular microscopy (Tomey Corporation, Nagoya, Japan), and slit-lamp examination of the anterior and posterior ocular segments.
All procedures were performed at Vardinoyiannion Eye Institute of Crete by the same surgeon (GDK) under sterile conditions. Apart from the different epithelial removal technique, the CXL procedure was the same in both groups.
After topical anesthesia with proxymetacaine hydrochloride 0.5% eye drops (Alcaine; Alcon Laboratories, Inc., Fort Worth, TX), the corneal epithelium was removed. Epithelial removal was performed by t-PTK using the PTK mode of the Allegretto WaveLight excimer laser (WaveLight Technologies, Erlangen, Germany). T-PTK ablation was performed in a 7-mm zone with an intended depth of 50 µm and then the deepithelialized area was enlarged mechanically by scraping the epithelium with a spatula (Tutton double-ended flat oval spatula/elevator; Duckworth & Kent Ltd., Baldock, United Kingdom) at an intended diameter of 8 mm. After epithelial removal, riboflavin (0.1% solution of 10 mg riboflavin-5-phosphate in 10 mL dextran-T-500 20% solution) was instilled on the center of the cornea every 3 minutes for 30 minutes. Saturation of the corneal stroma (riboflavin shielding) was monitored by the surgeon. UVA was performed using a UVA optical system (UV-X illumination system version 1000; IROC AG, Zurich, Switzerland) with a light source consisting of an array of ultraviolet diodes (370 nm) with a potentiometer in series to allow regulation of voltage. Before treatment, an intended irradiance of 3 mW/cm2 was calibrated using the UVA light meter YK-34UV (Lutron Electronic Enterprise Co., Ltd., Taipei, Taiwan), which is supplied with the UV-X device. Irradiance was performed for 30 minutes, corresponding to a total surface dose of 5.4 J/cm2. During UVA irradiation, riboflavin solution was applied every 3 minutes to maintain corneal saturation with riboflavin.
After topical anesthesia with proxymetacaine hydrochloride 0.5% eye drops (Alcaine; Alcon Laboratories, Inc.), the corneal epithelium was removed by mechanical debridement using a rotating brush at an intended diameter of 8 mm. After epithelial removal, the CXL procedure was followed as in the Cretan protocol group.
Table 1 summarizes the data of the CXL procedure for both groups. At the end of the treatment, a combination of antibiotic and corticosteroid (Tobradex; Alcon Laboratories, Inc.) drops was administered and a silicon-hydrogel bandage contact lens (Lotrafilcon B; Air Optix, Ciba Vision, Duluth, GA) was applied until full reepithelialization. A combination of antibiotic and corticosteroid drops was continued four times daily until the removal of the bandage contact lens. After the removal of the contact lens, patients received corticosteroid drops (FML, fluorometholone 0.1%; Falcon Pharmaceuticals, Fort Worth, TX) tapering for the next 3 weeks. Patients were encouraged to use artificial tears at least six times per day for 3 months postoperatively.
Microsoft Excel software (Microsoft Corporation, Redmond, WA) and SPSS version 18.0 software (SPSS, Inc., Chicago, IL) were used for data analysis. All data were examined for normality; because normality was not valid, parametric statistical analysis was performed for all parameters. All values are expressed as mean ± standard deviation and range. A P value of less than .05 was considered statistically significant. Visual acuity is expressed as logMAR units.
Follow-up time was 4 years for all patients included in the study. No intraoperative or postoperative complications were observed in any of the patients during the follow-up period. Tables A–B (available in the online version of this article) show the parameters evaluated preoperatively and at each postoperative interval (1, 2, 3, and 4 years postoperatively) for the Cretan and Dresden protocol groups, respectively. No statistically significant differences were found between the two groups preoperatively for any of the parameters evaluated after statistical examination using Wilcoxon/Kruskal–Wallis tests (rank-sums) (Tables A–B).
Preoperative and Postoperative Patient Data of t-PTK Group (Cretan Protocol)
Preoperative and Postoperative Patient Data of Mechanical Epithelial Debridement Group (Dresden Protocol)
Visual and Refractive Outcomes
In the Cretan protocol group, both mean UDVA and CDVA improved significantly (P < .05) from the first postoperative year and the significant improvement remained throughout the 4-year follow-up (Table A, Figure 1). By last follow-up, 60% of eyes gained one or more lines of CDVA (13.3% gained three or more lines, 6.7% gained two lines, and 40% gained one line of CDVA), 33.3% of eyes neither gained nor lost lines of CDVA, and 6.7% lost one line of CDVA (safety index = 1.28) (Figure 2A).
Change of (A) uncorrected (UDVA) and (B) corrected distance visual acuity (CDVA) in each group during the follow-up period. t-PTK = transepithelial phototherapeutic keratectomy; CXL = corneal cross-linking; preop = preoperatively
Change in corrected distance visual acuity (CDVA) bar graph (safety) in the (A) transepithelial phototherapeutic keratectomy (t-PTK) and (B) mechanical epithelial debridement groups. postop = postoperatively
In the Dresden protocol group, mean UDVA improved significantly (P < .05) from the first postoperative year and the significant improvement remained throughout the 4-year follow-up (Table B; Figure 1A). On the contrary, CDVA changed postoperatively but the improvement did not reach statistical significance (P > .05) at any of the postoperative time points (Table B; Figure 1B). By last follow-up, 40% of eyes gained one or more lines of CDVA (6.7% gained three or more lines, 20% gained two lines, and 13.3% gained one line of CDVA), 46.7% of eyes neither gained nor lost lines of CDVA, and 13.3% lost two lines of CDVA (safety index = 1.21) (Figure 2B).
Mean spherical equivalent in eyes of both groups did not change significantly (P > .05) at any postoperative interval (Tables A–B).
In the Cretan protocol group, mean steep and flat keratometry readings (D) improved significantly (P < .05) from the first postoperative year and the significant improvement remained throughout the 4-year follow-up (Table A). Mean corneal astigmatism also improved postoperatively and the improvement was significant (P < .05) from the second postoperative year (Table A).
In the Dresden protocol group, steep and flat keratometry readings improved postoperatively, but the improvement was significant (P < .05) after the second postoperative year (Table B). Mean corneal astigmatism changed postoperatively but the improvement did not attain significance (P > .05) at any postoperative interval (Table B).
Mean corneal thickness did not change significantly (P > .05) at any postoperative interval for both groups (Tables A–B).
Mean ECD did not change significantly (P > .05) at any postoperative interval for both groups (Tables A–B).
According to the Dresden protocol, corneal epithelium has to be removed during CXL because the efficacy of the procedure depends on adequate corneal stroma saturation with riboflavin.1 According to the Cretan protocol, epithelial removal with t-PTK during CXL results in better visual and refractive outcomes in patients with keratoconus compared to mechanical epithelial debridement.8,9 We reported significant visual and topographic improvement after t-PTK-CXL in a patient with keratoconus.10 Then, in a prospective comparative case series, we showed that there was a significant improvement in UDVA, CDVA, steep keratometry, and corneal astigmatism 12 months postoperatively in patients undergoing t-PTK epithelial removal during CXL, whereas in the mechanical epithelial debridement group there were no similar results.8 Afterward, we reported significant improvement in UDVA, CDVA, steep and flat keratometric values, and corneal astigmatism over a long-term follow-up after performing the Cretan protocol in a series of patients with keratoconus.11 However, there was no control group in this long-term prospective case series.11
In this article, we prospectively compared the visual, refractive, and keratometric long-term results of CXL in two groups of patients with progressive keratoconus using either t-PTK (Cretan protocol) or mechanical debridement (Dresden protocol) for epithelial removal. In accordance with the results of our previous studies, there was a significant improvement in visual acuity (both UDVA and CDVA) and keratometry readings (both steep and flat) in the Cretan protocol group from the first postoperative year, which remained significant throughout the follow-up period. Corneal astigmatism also improved significantly from the second postoperative year. In the Dresden protocol group, UDVA improved significantly but CDVA did not show significant improvement postoperatively. Both steep and flat keratometry readings improved significantly in the Dresden protocol group, but only after the second postoperative year, whereas corneal astigmatism did not improve significantly postoperatively. A possible explanation for the significant improvement in keratometry readings after the second postoperative year in the Dresden protocol group is the continuing long-term flattening effect of the cornea after CXL.3,4 Furthermore, 60% of eyes in the Cretan protocol group gained one or more lines of CDVA compared to 40% of the CXL eyes at 4 years postoperatively. None of the eyes in the Cretan protocol group lost two or more lines of CDVA, whereas 13.3% of the Dresden protocol group did at the 4-year postoperative interval.
The current study confirms our previous reports that t-PTK epithelial removal is advantageous in comparison with mechanical epithelial debridement during CXL, resulting in faster and significant improvement in visual, refractive, and keratometric values. As already described in our previous studies, these outcomes are due to the smoothing of the anterior corneal stroma caused by the t-PTK removal of the stromal tissue at the apex of the cone of the ectatic cornea.8,11 Some parameters, such as CDVA and corneal astigmatism, showed significant improvement only in the Cretan protocol group; other parameters, such as keratometry readings, showed significant improvement from the first postoperative year in the Cretan protocol group, whereas their improvement attained significance after long-term follow-up in the Dresden protocol group. Moreover, the percentage of eyes gained one or more lines of CDVA was significantly higher in the Cretan protocol group than in the Dresden protocol group at last follow-up. ECD remained unchanged in both groups during the 4-year follow-up.
Several other studies have also evaluated the combined t-PTK-CXL technique of the Cretan protocol by comparing t-PTK versus mechanical epithelial debridement during CXL.12,13 In a retrospective, comparative study, Kapasi et al.12 reported better visual outcomes in the t-PTK-CXL group than the CXL group 12 months after treatment. A significant improvement in CDVA and gain of CDVA lines were reported only in the Dresden protocol group 12 months postoperatively.12 In another retrospective comparative study, Gaster et al.13 indicated similar outcomes between t-PTK-CXL and mechanical epithelial debridement during CXL with a follow-up of 24 months. Both studies showed significant improvement of CDVA only in the t-PTK-CXL group at the last follow-up.12,13 Although Gaster et al. concluded that both t-PTK and mechanical debridement for epithelial removal during CXL yielded similar outcomes, CDVA improved significantly at 24 months only in the t-PTK-CXL group.13,14 Moreover, in the Gaster et al.'s study, the t-PTK-CXL group had reduced corneal steepness (significantly decreased maximum keratometry) compared to the manual epithelial debridement group at the 6-month postoperative interval.13 Although that difference was not demonstrated thereafter, it corroborates the basis of the combined t-PTK-CXL technique, according to which t-PTK removes a small amount of corneal stroma tissue at the cone apex, consistent with the epithelial doughnut pattern of Reinstein et al.,15 regularizing the anterior corneal surface and acting as a customized treatment during CXL.8,11,13 Higher values of maximum keratometry in the t-PTK-CXL group of Gaster et al. could have enhanced the postoperative effect of t-PTK with subsequent greater improvement in keratometric and visual values.13,14
Apart from the preoperative maximum and steep keratometric values of the keratoconic cornea, the preoperative epithelial thickness distribution seems to be another significant parameter that could also affect the effectiveness of the t-PTK-CXL procedure. The lack of epithelial thickness measurements is a limitation of our study and a significant parameter for the postoperative outcome of this combined technique that has not been evaluated until now. The customization of t-PTK-CXL and the t-PTK parameters (ablation depth and zone of transepithelial PTK) according to the preoperative epithelial map and corneal profile of patients with keratoconus could maximize the effect of this combined treatment. Broadly, t-PTK-CXL could be more effective in patients with more advanced keratoconus.
Our current long-term study indicates that t-PTK-CXL (Cretan protocol) results in better visual, keratometric, and refractive outcomes than CXL with mechanical epithelial debridement. Although other studies reported similar outcomes between the two methods of epithelial removal during CXL, all studies up to now support the conclusion that t-PTK-CXL is an effective and safe combined treatment for patients with keratoconus that not only stabilizes the ectatic cornea, but also significantly improves the vision of the patients postoperatively. In our opinion, this combined technique could be performed in any case of CXL aiming for better visual and refractive outcomes, especially cases in which photorefractive keratectomy cannot be combined with CXL due to low corneal thickness. Future prospective comparative studies taking into consideration several preoperative parameters, such as epithelial thickness distribution and keratometry values, and respective customization of t-PTK-CXL are of great importance to confirm the outcomes and evaluate the cases in which the effectiveness of this combined technique could be greater with minimal risk.