C orneal ectatic diseases such as keratoconus are progressive disorders in which stromal thinning and biomechanical weakening lead to irregular astigmatism and visual deterioration. The clinical burden of keratoconus is greater than historically appreciated, with many patients now presenting with advanced disease and very thin corneas at their initial evaluation. Preventing further progression is therefore essential to preserve visual function and avoid or delay keratoplasty.

Corneal cross-linking (CXL) exploits the photochemical interaction of riboflavin, oxygen, and ultraviolet (UV) light to generate reactive oxygen species (ROS) within the corneal stroma. ^,, These ROS induce covalent cross-links between stromal collagen and proteoglycan molecules, thereby increasing stromal rigidity. Over the past two decades, CXL has become the backbone of managing most corneal ectasias, helping many patients avoid keratoplasty. Notably, the original CXL “Dresden” protocol, developed over 20 years ago, has served as the gold-standard method for strengthening ectatic corneas. The Dresden protocol delivers a UV fluence of 5.4 J/cm² through 3 mW/cm² continuous irradiation over 30 minutes.

However, the Dresden protocol also specified a minimum corneal stromal thickness of 400 µm to ensure an uncross-linked stromal safety margin of approximately 70 µm above the endothelium. The rationale for this margin originates from experimental studies, predominantly in porcine ex-vivo corneas and rabbit in-vivo models, demonstrating endothelial susceptibility to riboflavin–UV exposure. ^, Direct transferability of these thresholds to humans in vivo has not been formally established; nevertheless, these findings have informed CXL safety principles for two decades, and we therefore retain the ∼70 µm uncross-linked safety margin as a conservative design constraint. This restriction excluded many clinically advanced ectatic corneas that could have benefited from treatment.

Several thin-cornea CXL approaches attempted to adapt the cornea to the Dresden protocol by artificially increasing preirradiation stromal thickness, including hypoosmolar riboflavin swelling, contact lens–assisted CXL (CACXL), and epithelial-island CXL, each with intrinsic limitations such as unpredictable swelling, restricted oxygen diffusion, or irregular energy distribution. These limitations motivated a shift toward adapting CXL parameters to the patient, rather than adapting the patient’s cornea to the protocol.

To address this, we previously developed and published a mathematical model describing UV–riboflavin–oxygen interactions within the stroma, enabling prediction of cross-linking depth based on stromal thickness. This model formed the basis of the first-generation ELZA-sub400 protocol, which individualized UV fluence delivery during continuous 3 mW/cm² irradiation to maintain an uncross-linked stromal safety margin of approximately 70 µm while effectively cross-linking thin corneas. This approach safely extended CXL to corneas with stromal thickness <400 µm, substantially broadening the therapeutic range.

However, many contemporary cases present with more aggressive disease, higher preoperative keratometry values, prior progression after CXL, or rapid deterioration, motivating the need for a more potent and efficient CXL strategy. Two refinements informed the second-generation approach. First, higher UV fluences—up to approximately 10 J/cm²—have been shown to produce biomechanical strengthening comparable to the Dresden protocol in accelerated regimens, without compromising endothelial safety in riboflavin-saturated stroma. ^, Second, higher irradiance settings (eg, 9 mW/cm² rather than 3 mW/cm²) allow more rapid delivery of a prescribed UV fluence. Importantly, the change here is in irradiance (rate of energy delivery, mW/cm²), not in total fluence (J/cm²): the kinetics of intrastromal oxygen replenishment are governed by irradiance, not by fluence, and we are not claiming that high fluence itself preserves oxygen.

Building on these insights, we evaluated a second-generation ELZA-sub400 protocol that maintains the individualized fluence scaling of the first-generation approach but incorporates two refinements: (1) an expanded total fluence ceiling of up to 10 J/cm², and (2) the incorporation of controlled accelerated UV irradiation at 3 mW/cm² or 9 mW/cm². The relative scaling of energy with intraoperative pachymetry, derived from our published nomogram, remains unchanged. The aim of this second-generation protocol is to deliver sufficient biomechanical strengthening for advanced or rapidly progressive ectatic disease while preserving endothelial safety through individualized energy titration.

This study presents the clinical outcomes of this second-generation ELZA-sub400 protocol in ultrathin ectatic corneas with stromal thicknesses below 400 µm.

This was a retrospective, single-center, consecutive case series. Surgeries were performed between June 2019 and May 2023 at the ELZA Institute in Dietikon/Zurich, Switzerland, and data were collected retrospectively from the institutional clinical database. The Cantonal Ethics Committee of the Canton of Zurich approved the retrospective use of these data (BASEC number 2018-0 2369). Written informed consent was obtained from all patients for both the treatment and the use of their de-identified data in research. The study adhered to the Declaration of Helsinki, the principles of Good Clinical Practice, the Swiss Human Research Act, the Swiss Human Research Ordinance, and all applicable local regulations. No clinical trial registration was required, as this was a retrospective observational analysis rather than a prospective clinical trial; no animals were involved.

Twenty-nine eyes of 24 patients with progressive keratoconus or post-LASIK ectasia and a post-soak intraoperative thinnest stromal thickness <400 µm were included. Patient sex was recorded in the medical record as male or female and is reported as such throughout; the term “sex” is used uniformly, and where cited prior literature used different terminology, the original wording has been preserved. Eligibility was defined a priori as keratoconus or post-LASIK ectasia meeting one or more of the following progression criteria: (1) an increase in anterior keratometry of ≥1 D within the previous 12 months; (2) historical cumulative anterior keratometric steepening of ≥4 D over ≤40 months; (3) documented progression despite previous CXL, or documented prior progression in the setting of topographic stability at the time of inclusion; (4) post-LASIK ectasia at high risk of progression, with or without documented 12-month change at the time of inclusion; or (5) in patients under the age of 20 years, primary ectasia in whom progression was presumed on the basis of age at presentation and who were therefore treated at first presentation. ^, Young age (<20 years) is itself an independent risk factor for progression in keratoconus and was treated as a high-risk criterion for inclusion. Anterior keratometric change was evaluated through differential maps comparing serial Scheimpflug tomography or Placido examinations; where differential maps were unavailable because of software incompatibility or a prior nonelectronic exam, maximum keratometry (Kmax) was used.

Exclusion criteria included: a history of >10 pack-years of tobacco smoking, ^,, pregnancy or lactation, pre-existing ocular trauma, inability to understand the nature of the study and/or give consent, and patients under guardianship. Previous ocular surgery was not an exclusion when it consisted of (1) prior laser vision correction in eyes presenting with post-LASIK ectasia, or (2) prior corneal cross-linking in eyes meeting the inclusion criteria for documented progression after CXL.

Clinical data, including corrected distance visual acuity (CDVA), refraction, and biomicroscopy, were recorded before surgery and postoperatively at 1 month and twelve months after CXL. To assess the depth of the demarcation line after CXL, anterior segment optical coherence tomography (AS-OCT) was performed during the 1-month postoperative consultation using spectral-domain OCT technology (Spectralis HRA version 1.10.0.0, Heidelberg Engineering, Heidelberg, Germany, or MS-39, CSO, Florence, Italy). The distances from the demarcation line to the anterior stroma and from the demarcation line to the endothelium were recorded. In all patients, such measurements were performed by the same examiner, and distances were measured at the thinnest point of the cornea.

All subjects had corneal evaluations performed using a rotational Scheimpflug system (Pentacam HR, Oculus, Wetzlar, Germany) by the same trained individual. The standard resolution setting was used to capture images (25 images per scan), and the following parameters were recorded: thinnest corneal stromal thickness, anterior radius of curvature in the 3.0 mm zone centered on the thinnest location of the cornea (ARC3mm), maximum anterior keratometry (K _max ), total anterior densitometry (AntDens), total central densitometry (CenDens), total posterior densitometry (PostDens) and total average densitometry (TotalDens). According to the Scheimpflug system’s standard parameters, AntDens corresponds to the 120 µm most superficial corneal layers, and the PostDens corresponds to the 60 µm closest to the endothelium.

THE SECOND-GENERATION ELZA-SUB400 PROTOCOL

Table 1 summarizes the technical specifications and CXL surgical principles. CXL was performed by mechanically removing the epithelium over 9 mm of the central cornea. Following de-epithelialization, the cornea was soaked with sodium edetate and trometamol–enriched riboflavin phosphate 0.1% hypotonic solution (Ribo-Ker; EMAGine AG, Zug, Switzerland) for 10 minutes. Ultrasound pachymetry was performed every 5 minutes during soaking to monitor eventual changes in corneal stromal thickness (Tomey, SP-1000, Nagoya, Japan). Guided by the preoperative Scheimpflug images, the intraoperative pachymetry measurements were performed in the thinnest area of the cornea. As a routine, ten measurements were taken in the thinnest area, and the lowest value was considered. At the end of the soaking period, corneas were rinsed with balanced salt solution to rinse off any surplus of riboflavin, and ultrasound pachymetry was performed to determine minimal stromal thickness. This intraoperative pachymetry measurement was performed at the end of the riboflavin instillation, as this post-soaking minimum corneal thickness value was required to determine the patient’s individualized fluence requirement, per our published nomogram, aiming to obtain a demarcation line 70 µm above the corneal endothelium.

TABLE 1

Technical Specifications Used in the Second-Generation ELZA-sub400 Protocol.

Parameter	Individualized CXL
Treatment target	Keratoconus/ iatrogenic ectasia
Fluence (total) (J/cm²)	Variable (10 J/cm², fractionated by thinnest point corneal pachymetry)
Soak Time (minutes) & interval	10 (q2)
Intensity (mW/cm²)	3 or 9
Treatment time (minutes)	Variable
Epithelium status	Off
Chromophore	0.1% riboflavin
Light Source	C-eye
Irradiation mode	Continuous

Crucially, the post-soak intraoperative thinnest stromal thickness was the variable that determined the individualized UV fluence to be delivered (in 10 µm increments, per the published nomogram10), with the goal of placing the demarcation line approximately 70 µm above the endothelium ( Table 2 ). The choice of irradiance (3 mW/cm² or 9 mW/cm²) did not change the prescribed fluence; it changed only the irradiation time required to deliver that fluence. Total UV fluence was up to 10 J/cm², with a maximum total UV exposure time of approximately 18 minutes at 9 mW/cm² (vs approximately 30 minutes at 3 mW/cm² in the first-generation protocol). UV irradiation was delivered at 365 nm using a commercial CXL device (C-eye; EMAGine AG, Zug, Switzerland).

TABLE 2

Individualized CXL Parameters by Post-Soak Intraoperative Thinnest Stromal Thickness, Scaled in 10 µm Increments and Targeting a 70 µm Uncross-Linked Safety Margin Above the Endothelium.

Min Required Thickness (µm)	Intensity (mW/cm²)	UV Irradiation Duration (min)	Demarcation Line (µm)
200	3	1:51	130
210	3	2:28	140
220	3	3:04	150
230	3	3:42	160
240	3	4:38	170
250	3	5:33	180
260	3	6:29	190
270	3	7:24	200
280	3	9:15	210
290	3	11:06	220
300	3	12:57	230
310	3	16:39	250
320	3	18:30	255
330	9	7:24	265
340	9	8:38	275
350	9	9:52	283
360	9	11:06	290
370	9	12:20	300
380	9	14:11	310
390	9	16:02	320
400	9	17:53	330

Note: choice of irradiance (3 vs 9 mW/cm²) determines the irradiation time required to deliver the prescribed fluence; it does not change the fluence itself.

Postoperatively, a bandage contact lens was placed immediately after surgery, and antibiotic and corticosteroid drops were administered. Patients were re-examined at the slit lamp on postoperative day 1, and daily until the epithelium was closed, as well as after 1 month and twelve months after the procedure. The contact lens was removed usually on day 4. Postoperative drops included fourth-generation fluoroquinolone antibiotics twice a day for 7 days, followed by 0.1% fluorometholone drops twice a day for 12 weeks, and preservative-free artificial tears as needed.

Only gold members can continue reading. Log In or Register to continue

Share this:

• X
• Facebook

Related

Related posts:

1. Zeb1 Mediates Injury-Induced Mesenchymal Transition in the Mouse Corneal Endothelium In Vivo
2. Associations Between Routine Vaccinations and the Risk of New-Onset Idiopathic Uveitis
3. Clinical Utility of Ultra-Widefield Swept-Source OCT for Intraocular Tumors: Comparison With Ultrasonography, SD-OCT, and MRI
4. Comment on: Visual Field Progression in Patients With Glaucoma With Delayed Follow-Up

Stay updated, free articles. Join our Telegram channel

Join