Cross-Linking for Ectasia
Corneal cross-linking (CXL) is a photochemical treatment that strengthens the cornea by inducing the formation of covalent bonds between collagen fibers and proteoglycans in the structural layer of the cornea, the stroma. This is achieved through the combined use of riboflavin (vitamin B2) and ultraviolet A (UV-A) light, which causes a photochemical reaction, generating reactive oxygen species, that then cross-link the stroma. The procedure remains the only intervention proven to halt keratoconus progression, and its widespread adoption has led to a 25% reduction in the need for corneal transplantation in some regions.
The Dresden protocol is the original and most validated corneal cross-linking method, delivering 5.4 J/cm² of total UV fluence over 30 minutes at a UV intensity of 3 mW/cm². The epithelium is removed to allow riboflavin penetration into the stroma, ensuring effective cross-link formation (riboflavin is too big to penetrate through the epithelial cell-cell tight junctions; epithelium removal makes this an “epi-off” technique).
While highly effective, the Dresden protocol has several drawbacks. The procedure is lengthy, requiring 30 minutes of UV exposure, which limits clinical efficiency. Epithelial removal exposes the patient to postoperative discomfort and infection risk, necessitating careful pain management. Additionally, the protocol is restricted to corneas with a minimum stromal thickness of 400 µm, as thinner corneas are at risk of UV-induced endothelial damage, limiting its applicability in advanced keratoconus cases.
How have things changed since CXL’s introduction over a quarter of a century ago?
Accelerating CXL should have been easy – if it weren’t for one little thing.
The Bunsen-Roscoe law of reciprocity states that for photochemical reactions, the total energy (fluence) delivered determines the reaction outcome, regardless of whether it is applied over a longer period at lower intensity or a shorter period at higher intensity. This principle led to the development of accelerated CXL protocols, where higher UV intensity was used to shorten treatment time while maintaining the same total fluence as the Dresden protocol.
However, early studies observed diminishing biomechanical returns with increasing UV intensity, indicating that simply reducing treatment time did not achieve the same stiffening effect. This was eventually found to be due to oxygen depletion in the stroma, as higher UV intensities consume oxygen faster than it can diffuse from the atmosphere, limiting the formation of reactive oxygen species (ROS) required for effective cross-linking.
| Protocol | UV Fluence (J/cm²) | UV Intensity (mW/cm²) | Treatment Time (min) | Biomechanical Considerations |
| Dresden Protocol | 5.4 | 3 | 30 | High efficacy, stromal oxygen maintained |
| 10-Min Protocol | 5.4 | 9 | 10 | Reduced stromal oxygen availability |
| 5-Min Protocol | 5.4 | 18 | 5 | Stromal oxygen depletion limits effect |
| High-Fluence (10J/cm²) | 10.0 | 18 | 9:15 | Increased energy compensates for oxygen limitations |
The original Dresden protocol delivers a total UV fluence of 5.4 J/cm² over 30 minutes at an intensity of 3 mW/cm². The long exposure time allows continuous diffusion of atmospheric oxygen into the stroma, ensuring sufficient reactive oxygen species (ROS) generation to facilitate effective cross-linking. The stiffening effect extends to approximately 330 µm, with a 70 µm safety margin above the corneal endothelium.
This protocol reduces treatment time by increasing UV intensity. However, the higher UV intensity leads to faster oxygen consumption than diffusion can replenish, partially reducing the biomechanical stiffening effect. Pulsed UV delivery has been introduced to mitigate this limitation by allowing oxygen levels to recover intermittently.
This protocol maintains the same fluence as Dresden but delivers it in five minutes using 18 mW/cm² intensity. The rapid oxygen depletion in the stroma limits the cross-linking effect, reducing overall biomechanical strengthening unless oxygen supplementation is used.
This protocol delivers a higher total fluence of 10 J/cm² while maintaining a UV intensity of 18 mW/cm². The increased fluence partially compensates for oxygen depletion, producing a biomechanical effect similar to Dresden but in a shorter timeframe.
Pulsed UV-A irradiation and external oxygen supplementation have been introduced to enhance oxygen availability and improve biomechanical outcomes in accelerated protocols.
The primary limitation of the Dresden protocol is its requirement for a minimum corneal stromal thickness of 400 µm to prevent UV-induced endothelial damage. Many patients with advanced keratoconus fall below this threshold, necessitating alternative approaches to enable safe cross-linking in thin corneas. Several strategies have been developed to address this challenge, each with different limitations and biomechanical outcomes.
Hypoosmolar riboflavin solution is used to induce transient corneal swelling, artificially increasing stromal thickness above the 400 µm threshold before UV irradiation. However, corneal dehydration during the procedure can result in thickness reduction below safe levels, limiting its predictability.
A riboflavin-soaked contact lens is placed on the cornea to serve as a protective layer during UV irradiation. This method enables treatment of corneas as thin as 330 µm, but the presence of the contact lens reduces oxygen diffusion to the stroma, leading to diminished biomechanical stiffening.
The M protocol applies varying UV intensities and durations to achieve an individualized cross-linking depth, maintaining a 70 µm endothelial safety margin. This approach improves safety but requires access to multiple CXL devices capable of delivering customized treatment settings – essentially one combination of reagents and devices per required depth of cross-linking effect.
The ELZA-sub400 protocol was developed to allow individualized CXL treatment in corneas as thin as 214 µm. Using an algorithm that models the interaction between UV, riboflavin, stromal tissue and oxygen, this approach calculates the appropriate UV fluence and duration based on post-epithelial removal pachymetry measurements, ensuring the treatment remains within the ~70 µm uncross-linked endothelial safety margin. The ELZA-sub400 protocol is device-independent, as all CXL platforms can deliver the required UV intensities. This method represents the most straightforward approach for safely performing CXL in thin and ultra-thin corneas.
The choice of protocol depends on multiple factors, including disease severity, corneal thickness, and procedural constraints. While accelerated protocols reduce treatment time, they may result in reduced biomechanical stiffening if oxygen levels are not optimized.
The Dresden protocol remains the reference standard due to its validated long-term efficacy.
10-minute and 5-minute protocols shorten treatment duration but require oxygen management strategies to maintain their biomechanical effect.
High-fluence CXL enables safe treatment in thin corneas and rapidly progressing disease by increasing total fluence while compensating for oxygen limitations.
Thin cornea protocols, including hypoosmolar riboflavin, CA-CXL, the M protocol, and ELZA-sub400, expand the indications for CXL but vary in efficacy and predictability.
Epi-off CXL protocols differ in biomechanical outcomes due to variations in UV intensity, treatment duration, and stromal oxygen availability. The Dresden protocol remains the most validated approach, but accelerated, high-fluence, and customized techniques provide alternatives for patients with thinner corneas or procedural time constraints. Ongoing research has continued to refine these protocols to optimize safety and efficacy, and we are now at a point where accelerated, high-fluence protocols can deliver Dresden protocol-like cross-linking strengthening effects, in a considerably shorter period.
