One‐Year Results of Accelerated Transepithelial Corneal Cross‐Linking for Progressive Keratoconus: A Prospective Interventional Study

Abstract

Background: Keratoconus is a progressive corneal ectatic disorder. Accelerated transepithelial corneal cross‐linking (ATE‐CXL) aims to halt progression while preserving the epithelium, but evidence on its 12‐month efficacy requires further evaluation.

Purpose: To evaluate the 12‐month topographic, visual, biomechanical, and safety outcomes of ATE‐CXL in eyes with progressive keratoconus.

Method: This prospective interventional study included 126 eyes from 63 patients with documented progressive keratoconus. Both eyes of each patient were included when the eligibility criteria were met. Progression before treatment was defined as at least one of the following during the preceding 12 months: an increase in maximum keratometry (Kmax) of ≥ 1.0 diopter (D), a decrease in minimum corneal thickness of > 10 μm, or a loss of more than two lines of corrected distance visual acuity (CDVA). All eyes underwent epithelium‐on ATE‐CXL using a two‐step transepithelial riboflavin protocol followed by pulsed ultraviolet‐A irradiation at 45 mW/cm² for 320 s, delivering 7.2 J/cm². Outcomes included Kmax, flat keratometry (K1), steep keratometry (K2), uncorrected distance visual acuity (UDVA), CDVA, corneal hysteresis (CH), corneal resistance factor (CRF), central corneal thickness (CCT), progression, and adverse events. Outcomes were summarized descriptively using aggregate values.

Results: Kmax decreased from 52.3 ± 1.8 D at baseline to 49.7 ± 1.5 D at 12 months. CDVA improved from 0.35 ± 0.10 logMAR to 0.20 ± 0.08 logMAR, and UDVA improved by 30.9% from baseline. CH increased from 8.4 ± 1.0 mmHg to 9.1 ± 1.2 mmHg, and CRF increased from 8.0 ± 1.1 mmHg to 8.8 ± 1.3 mmHg. Disease progression, defined as Kmax increase ≥ 1.0 D at 12 months, occurred in 8 of 126 eyes (6.3%). Accordingly, 118 of 126 eyes (93.7%) remained stable by the prespecified Kmax criterion. No severe adverse events were documented.

Conclusion: ATE‐CXL was associated with 12‐month improvements in corneal topography, visual acuity, and ocular‐response‐analyzer biomechanical parameters in this prospective cohort of eyes with progressive keratoconus. These findings should be interpreted cautiously because the study lacked an epithelium‐off or untreated control group, had only 12 months of follow‐up, and included both eyes from each patient. Longer‐term controlled studies with paired‐eye‐adjusted statistical analysis are needed to confirm durability and comparative efficacy.

Graphical abstract

Keywords

accelerated transepithelial corneal cross‐linking / biomechanics / corneal stability / keratoconus / riboflavin / ultraviolet‐A / visual acuity

1 Introduction

Keratoconus is a progressive, usually bilateral, non‐inflammatory corneal ectatic disorder characterized by stromal thinning, conical protrusion, irregular astigmatism, and visual impairment [1]. The disease typically begins around puberty or adolescence and often progresses during the second and third decades of life, although onset and progression vary considerably among patients [1]. Contemporary epidemiological studies suggest that keratoconus may be more common than historical estimates, partly because of improved detection using corneal tomography [2].

Corneal collagen cross‐linking (CXL) is the principal intervention designed to slow or halt biomechanical progression in keratoconus [3]. The standard epithelium‐off Dresden protocol has demonstrated long‐term effectiveness [4]. However, epithelial removal is associated with postoperative pain, delayed epithelial healing, temporary visual recovery, and epithelial‐healing‐related complications [5, 6]. These limitations have encouraged the development of accelerated and transepithelial approaches intended to reduce treatment burden while preserving corneal stability [7–9].

Accelerated CXL protocols increase ultraviolet‐A (UVA) irradiance while reducing exposure time, broadly based on the Bunsen–Roscoe principle of reciprocity [7, 9]. Transepithelial, or epithelium‐on, CXL preserves the corneal epithelium and may improve postoperative comfort and reduce epithelial‐healing‐related adverse events [8, 10]. Nevertheless, the intact epithelium restricts stromal riboflavin diffusion, which remains a major limitation of transepithelial techniques [10]. To address this issue, modified riboflavin formulations containing epithelial permeability enhancers and pulsed UVA delivery have been introduced to improve stromal riboflavin availability and oxygen‐dependent photochemical efficiency [4, 11].

The present study evaluated the one‐year topographic, visual, biomechanical, and safety outcomes of a pulsed high‐fluence accelerated transepithelial corneal cross‐linking (ATE‐CXL) protocol in eyes with documented progressive keratoconus, building on prior work evaluating accelerated and transepithelial CXL approaches [8–10, 12, 13]. The study specifically assessed maximum keratometry (Kmax), flat keratometry (K1), steep keratometry (K2), uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), corneal hysteresis (CH), corneal resistance factor (CRF), central corneal thickness (CCT), disease progression, and adverse events over 12 months.

2 Methods

2.1 Study Design and Ethics

This prospective interventional study was approved by the ethics board of the College of Medicine at the University of Fallujah, Iraq (approval No. 113‐2023, approved on January 12, 2023). It was conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from all participants after an explanation of the procedure and its potential risks and benefits. For patients younger than 18 years, written informed consent was obtained from parents or legal guardians, and patient assent was documented where applicable.

2.2 Patient Selection and Eligibility Criteria

A total of 126 eyes from 63 patients aged 15–40 years with documented progressive keratoconus were included. Both eyes of each enrolled patient were included when both met the eligibility criteria. Progressive keratoconus was defined as at least one of the following changes during the preceding 12 months: an increase in Kmax of ≥ 1.0 D, a decrease in minimum corneal thickness of > 10 μm, or a loss of more than two lines of CDVA.

Patients were excluded if they had corneal scarring or opacity interfering with imaging or visual assessment, previous ocular surgery, active ocular surface infection or inflammation, systemic connective tissue disease, pregnancy, or lactation. Contact lenses were discontinued for a minimum of 2 weeks before baseline imaging to allow corneal surface stabilization. Patients with CCT less than 400 μm were excluded from treatment.

2.3 Intervention Protocol

All procedures were performed under topical anesthesia using an epithelium‐on technique. The corneal epithelium was not intentionally removed. After placement of a sterile lid speculum, transepithelial riboflavin loading was performed using a two‐step ParaCel protocol (Avedro Inc.). ParaCel Part One, containing 0.25% riboflavin with hydroxypropyl methylcellulose, benzalkonium chloride, ethylenediaminetetraacetic acid (EDTA), and tris(hydroxymethyl)aminomethane (TRIS), was instilled at 60‐s intervals for 4 min to facilitate epithelial permeability. This was followed by ParaCel Part Two, containing 0.22% riboflavin in isotonic saline, instilled at 90‐s intervals for 6 min. The total riboflavin loading time was approximately 10 min.

After riboflavin loading, the ocular surface was rinsed with balanced salt solution (BSS). Pulsed UVA irradiation was delivered using the KXL System (Avedro Inc.) at an irradiance of 45 mW/cm² for 5 min and 20 s, using a 1‐s‐on/1‐s‐off pulsed mode, for a total energy dose of 7.2 J/cm², consistent with the broader accelerated CXL literature [4, 7, 9]. After irradiation, the cornea was rinsed again with BSS.

2.4 Postoperative Management

Postoperative therapy consisted of preservative‐free lubricants four times daily for 2 weeks, followed by tapering as clinically indicated. Topical fluorometholone 0.1% was instilled four times daily for 1 week, followed by tapering. A bandage contact lens was not routinely applied because the epithelium was preserved. Topical antibiotic prophylaxis was not routinely used. However, when clinically indicated, moxifloxacin 0.5% was prescribed four times daily for 1 week.

2.5 Follow‐Up and Outcome Measures

Patients were evaluated at baseline, day 1, and 1, 3, 6, and 12 months after treatment. Topographic and tomographic parameters included Kmax, K1, K2, CCT, and minimum corneal thickness, measured using Scheimpflug imaging (Pentacam HR, Oculus Optikgeräte GmbH). Visual outcomes included UDVA and CDVA, recorded in logMAR units. Biomechanical outcomes included CH and CRF, measured using the Ocular Response Analyzer (Reichert Technologies Inc., USA). Safety outcomes included epithelial disruption, corneal haze grade, infectious keratitis, corneal melt, visually significant scarring, endothelial complications, loss of two or more lines of CDVA, and need for retreatment. Progression at 12 months was defined as an increase in Kmax of ≥ 1.0 D from baseline. Patient satisfaction was assessed using a standardized 10‐point visual analog scale (VAS) at baseline and 12 months, with higher scores indicating greater satisfaction.

2.6 Statistical Analysis

Data were analyzed using SPSS version 25.0 (IBM Corp.). Continuous variables were summarized as mean ± standard deviation, and categorical variables were summarized as counts and percentages. Descriptive statistics were calculated for all outcome measures at each follow‐up visit. Because individual‐level paired data were unavailable for independent reanalysis, outcomes were summarized descriptively using aggregate values. No repeated‐measures modeling, adjusted pairwise testing, or formal subgroup inference was performed. These findings should be interpreted as descriptive aggregate outcomes rather than independently verified inferential statistical findings.

3 Results

3.1 Topographic and Visual Outcomes

Topographic parameters showed numerical improvement during the 12‐month follow‐up period (Table 1, Figure 1). Kmax decreased from 52.3 ± 1.8 D at baseline to 49.7 ± 1.5 D at 12 months, representing a mean reduction of 2.6 D. Flat keratometry (K1) decreased from a baseline mean of 44.8 ± 2.1 D to 42.9 ± 1.9 D at 12 months (mean reduction 1.9 D). Steep keratometry (K2) decreased from 54.2 ± 2.3 D at baseline to 52.1 ± 2.0 D at 12 months (mean reduction 2.1 D).

Visual acuity showed improvement during follow‐up (Table 2, Figure 2). CDVA improved from 0.35 ± 0.10 logMAR at baseline to 0.20 ± 0.08 logMAR at 12 months, corresponding to an approximate gain of 1.5 Snellen lines. UDVA improved by 30.9% from baseline. Among the 126 eyes, 89 eyes (70.6%) gained one or more lines of CDVA, while 28 eyes (22.2%) remained unchanged and 9 eyes (7.1%) lost one or more lines. No eyes lost two or more lines of CDVA.

3.2 Biomechanical Outcomes and Safety

Corneal biomechanical parameters showed numerical increases after treatment (Table 3, Figure 3). CH increased from 8.4 ± 1.0 mmHg at baseline to 9.1 ± 1.2 mmHg at 12 months, representing a mean increase of 0.7 mmHg. CRF increased from 8.0 ± 1.1 mmHg to 8.8 ± 1.3 mmHg at 12 months, representing a mean increase of 0.8 mmHg. As shown in Table 1, CCT decreased slightly from 456 ± 18 μm at baseline to 451 ± 16 μm at 12 months (mean reduction: 5 μm). As shown in Table 3, intraocular pressure (IOP) remained stable, with no significant changes observed throughout the follow‐up period.

No severe adverse events were documented during follow‐up (Table 4). Mild epithelial disruption occurred in 10 eyes (7.9%) within the first 24 h and resolved without sequelae by the 1‐week follow‐up. Transient grade 1 corneal haze occurred in six eyes (4.8%) and resolved spontaneously by 3 months. No cases of infectious keratitis, corneal melt, visually significant scarring, endothelial complications, or other severe adverse events were reported. One eye (0.8%) required retreatment due to continued progression at 6 months.

3.3 Treatment Stability and Progression

Disease progression, defined as a Kmax increase of ≥ 1.0 D at 12 months, occurred in 8 of 126 eyes (6.3%). Therefore, 118 of 126 eyes (93.7%) remained stable by the prespecified Kmax criterion. Progression occurred primarily among younger patients and eyes with thinner corneas. The mean age of eyes with progression was 18.3 ± 2.1 years, compared with 26.4 ± 7.8 years for stable eyes. The mean pachymetry in the progression group was 442 ± 12 μm, compared with 461 ± 15 μm in the stable group. These subgroup observations should be interpreted as descriptive.

3.4 Patient‐Reported Outcomes

Patient‐reported satisfaction increased from 70% at baseline to 88% at 12 months, as measured by a standardized 10‐point VAS. Reduced glare sensitivity was reported by 76% of patients, and improved functional vision was reported by 82% of patients. Mean satisfaction scores improved from 6.8 ± 1.9 at baseline to 8.3 ± 1.4 at 12 months.

4 Discussion

This prospective interventional study found that ATE‐CXL was associated with improvements in corneal topographic, visual, and ocular‐response‐analyzer biomechanical parameters over 12 months in eyes with progressive keratoconus. The procedure was not associated with severe adverse events in the observed cohort. These findings suggest that the investigated transepithelial accelerated protocol may be clinically useful for selected patients, but the absence of an epithelium‐off or untreated control group prevents direct conclusions regarding comparative efficacy.

The observed mean Kmax reduction of 2.6 D appears larger than that reported in many transepithelial CXL studies and should therefore be interpreted cautiously [8, 10–12]. Several factors may explain this magnitude, including baseline disease characteristics, measurement variability, regression to the mean, patient selection, treatment protocol differences, and epithelial or stromal remodeling. The study population consisted of relatively young patients (mean age 26.4 years) with documented progressive disease, which may represent a more aggressive phenotype than some published cohorts. Additionally, the pulsed high‐fluence protocol (45 mW/cm² for 320 s) may deliver greater stromal riboflavin saturation compared with some continuous‐wave transepithelial approaches. These findings should be interpreted as descriptive aggregate outcomes rather than independently verified inferential statistical findings.

The visual acuity improvements observed in this cohort are clinically relevant, particularly given the stable tomography and absence of sight‐threatening complications. Visual, refractive, and tomographic outcomes remain central endpoints in CXL studies [9]. However, visual improvement after CXL may result from multiple mechanisms, including corneal flattening, epithelial remodeling, changes in higher‐order aberrations, contact‐lens or spectacle changes, and measurement variability. The improvements reported here should not be attributed solely to biomechanical strengthening, but rather represent the combined effect of multiple optical and structural changes.

The increases in CH and CRF suggest improvement in ocular‐response‐analyzer biomechanical parameters after treatment. Nevertheless, biomechanical metrics derived from air‐puff deformation provide only an indirect proxy for true tissue stiffness [14–16]. Although the observed numerical increases are consistent with the intended therapeutic effect of CXL, they should be interpreted as surrogate markers rather than direct measures of stromal collagen cross‐linking density.

Compared with the standard epithelium‐off Dresden protocol, transepithelial ATE‐CXL has potential practical advantages, including epithelial preservation, less epithelial‐healing‐related morbidity, and faster early recovery [8, 10]. Conversely, epithelium‐off CXL remains the better‐established approach with stronger long‐term evidence [3, 4]. Because this study did not include a randomized epithelium‐off control group, the results should not be interpreted as demonstrating superiority, equivalence, or non‐inferiority compared with the Dresden protocol. The appropriate conclusion is that the investigated protocol produced favorable 12‐month outcomes in this cohort and warrants further controlled evaluation.

Progression occurred in 8 eyes (6.3%), mainly among younger patients and eyes with thinner corneas. This observation is clinically plausible because younger age and more advanced ectatic disease are commonly associated with higher progression risk [1, 2]. However, the present subgroup findings should be considered exploratory. Patients at higher risk, particularly those younger than 20 years or with CCT less than 450 μm, may require closer follow‐up, individualized treatment selection, or consideration of established epithelium‐off protocols when clinically appropriate.

5 Limitations

This study has several important limitations that should be considered when interpreting the results. First, it lacked an untreated, sham‐treated, or epithelium‐off control group, which limits causal inference and prevents direct comparison with the standard Dresden protocol. The observed improvements cannot be definitively attributed to the treatment, as some degree of natural variation or regression to the mean may have contributed. Second, the 12‐month follow‐up period is adequate for early assessment but insufficient to establish long‐term stabilization in a chronic progressive disease. Keratoconus typically progresses over years to decades, and longer follow‐up is needed to confirm sustained stability. Third, because individual‐level paired data were unavailable for independent reanalysis, no repeated‐measures modeling, adjusted pairwise testing, or confidence interval estimation was performed. The reported aggregate statistics should not be interpreted as independently verified inferential findings. Fourth, selection bias may have occurred because the cohort included patients considered suitable for transepithelial treatment, and the findings may not generalize to very advanced keratoconus, severe corneal thinning, corneal scarring, or pediatric patients at particularly high risk of progression. Fifth, the study included both eyes from each patient, which may introduce inter‐eye correlation that could affect the precision of estimates if not properly accounted for in statistical analysis. Finally, patient satisfaction was measured using a VAS, and the questionnaire was not formally validated in this population.

6 Conclusion

ATE‐CXL using a pulsed high‐fluence UVA protocol was associated with improvements in topographic, visual, and ocular‐response‐analyzer biomechanical parameters over 12 months in this prospective cohort of eyes with progressive keratoconus, consistent with the rationale for accelerated and transepithelial approaches but requiring confirmation against established epithelium‐off protocols. Disease progression occurred in 6.3% of eyes, while 93.7% remained stable by the prespecified Kmax criterion. No severe adverse events were documented. These findings support further evaluation of this minimally invasive protocol, but conclusions should remain conservative because of the absence of a control group, the limited follow‐up duration, and the inclusion of both eyes from each patient. Longer‐term controlled studies using paired‐eye‐adjusted analyses and comparison with epithelium‐off protocols are needed to clarify durability, safety, and comparative efficacy in diverse patient populations.

References

[1] J. A. Gomes, D. Tan, C. J. Rapuano, et al., “Global Consensus on Keratoconus and Ectatic Diseases,” Cornea 34, no. 4 (2015): 359–369.

[2] H. Hashemi, S. Heydarian, E. Hooshmand, et al., “The Prevalence and Risk Factors for Keratoconus: A Systematic Review and Meta‐Analysis,” Cornea 39, no. 2 (2020): 263–270.

[3] G. Wollensak, E. Spoerl, and T. Seiler, “Riboflavin/Ultraviolet‐A‐Induced Collagen Crosslinking for the Treatment of Keratoconus,” American Journal of Ophthalmology 135, no. 5 (2003): 620–627.

[4] G. Wollensak and H. Herbst, “Significance of the Lacunar Hydration Pattern After Corneal Cross Linking,” Cornea 29, no. 8 (2010): 899–903.

[5] P. Sarma, H. Kaur, F. Hafezi, et al., “Short‐ and Long‐Term Safety and Efficacy of Corneal Collagen Cross‐Linking in Progressive Keratoconus: A Systematic Review and Meta‐Analysis of Randomized Controlled Trials,” Taiwan Journal of Ophthalmology 13, no. 2 (2022): 191–202.

[6] G. Scarcelli, S. Besner, R. Pineda, and S. H. Yun, “Biomechanical Characterization of Keratoconus Corneas Ex Vivo With Brillouin Microscopy,” Investigative Ophthalmology & Visual Science 55, no. 7 (2014): 4490–4495.

[7] P. Vinciguerra, J. B. Randleman, V. Romano, et al., “Transepithelial Iontophoresis Corneal Collagen Cross‐Linking for Progressive Keratoconus: Initial Clinical Outcomes,” Journal of Refractive Surgery 30, no. 11 (2014): 746–753.

[8] M. A. Henriquez, G. Hernandez‐Sahagun, J. Camargo, and L. Izquierdo, “Accelerated Epi‐On Versus Standard Epi‐Off Corneal Collagen Cross‐Linking for Progressive Keratoconus in Pediatric Patients: Five Years of Follow‐Up,” Cornea 39, no. 12 (2020): 1493–1498.

[9] L. Lim and E. W. L. Lim, “A Review of Corneal Collagen Cross‐Linking: Current Trends in Practice Applications,” Open Ophthalmology Journal 12, no. 1 (2018): 181–213.

[10] X. Zhang, J. Zhao, M. Li, M. Tian, Y. Shen, and X. Zhou, “Conventional and Transepithelial Corneal Cross‐Linking for Patients With Keratoconus,” PLoS One 13, no. 4 (2018): e0195105.

[11] A. Caporossi, C. Mazzotta, S. Baiocchi, and T. Caporossi, “Long‐Term Results of Riboflavin Ultraviolet A Corneal Collagen Cross‐Linking for Keratoconus in Italy: The Siena Eye Cross Study,” American Journal of Ophthalmology 149, no. 4 (2010): 585–593.

[12] A. Caporossi, C. Mazzotta, A. L. Paradiso, S. Baiocchi, D. Marigliani, and T. Caporossi, “Transepithelial Corneal Collagen Crosslinking for Progressive Keratoconus: 24‐Month Clinical Results,” Journal of Cataract & Refractive Surgery 39, no. 8 (2013): 1157–1163.

[13] A. J. Kanellopoulos and G. Asimellis, “Revisiting Keratoconus Diagnosis and Progression Classification Based on Evaluation of Corneal Asymmetry Indices, Derived From Scheimpflug Imaging in Keratoconic and Suspect Cases,” Clinical Ophthalmology 7 (2013): 1539–1548.

[14] F. Raiskup, A. Theuring, L. E. Pillunat, and E. Spoerl, “Corneal Collagen Crosslinking With Riboflavin and Ultraviolet‐A Light in Progressive Keratoconus: Ten‐Year Results,” Journal of Cataract & Refractive Surgery 41, no. 1 (2015): 41–46.

[15] I. M. Beshtawi, R. Akhtar, M. C Hillarby, et al., “Biomechanical Changes After Repeated Collagen Cross‐Linking on Human Corneas Assessed in Vitro Using Scanning Acoustic Microscopy,” Investigative Ophthalmology & Visual Science 55, no. 3 (2014): 1549–1554.

[16] T. Almubrad, R. Mencucci, A. Smedowski, et al., “Ultrastructural Study of Collagen Fibrils, Proteoglycans and Lamellae of the Cornea Treated With Iontophoresis–UVA Cross‐Linking and Hypotonic Riboflavin Solutions,” Saudi Journal of Biological Sciences 28, no. 12 (2021): 7160–7174.

Rights & permissions

2026 The Author(s). Eye & ENT Research published by John Wiley & Sons Australia, Ltd on behalf of Higher Education Press.