TABLE 1.

Summary of existing corneal cross-linking protocols.

Protocol	Riboflavin delivery	Ultraviolet-A	Oxygen delivery	Efficacy
Dresden protocol	Epithelium-off	3 mW/cm2; 30 min	Room air	Sustained clinical outcomes up to 10 years postoperatively. (Raiskup and Spoerl, 2013; Wittig-Silva et al., 2014; Seyedian et al., 2015; Taşçı et al., 2020)
Ultraviolet-A irradiation
Accelerated protocol	Epithelium-off	Variable; 30 mW/cm2 for 3 min, 18 mW/cm2 for 5 min or 9 mW/cm2 for 10 minute	Room air	Regimens of low ultraviolet-A illumination with longer duration generally provide better results. (Peyman et al., 2016; Choi et al., 2017; Kirgiz et al., 2019) 9 mW/cm2 for 10 minute regimens are safe and efficacious in stabilizing keratoconus for up to 5 years follow-up (Mazzotta et al., 2021)
Ultraviolet-A-Emitting device	Epithelium-on 3 h/day for 6 months	0.31 mW/cm2 for 180 min daily for 6 months	Room air	Ex-vivo experiments on rabbit corneas have demonstrated that treated corneas with KeraVio were significantly stronger than the control group. Clinically, KeraVio halted corneal ectasia progression without any safety concerns in 20 eyes (Kobashi et al., 2020).
Riboflavin delivery
Transepithelial protocol	Epithelium-on	3 mW/cm2; 30 min/variable	Room air	Less effective than Dresden protocol. (Soeters et al., 2015)
Transepithelial protocol with: (1) chemical enhancers	Epithelium-on; (1) Loosening of epithelial tight junctions	3 mW/cm2; 30 min/variable	Room air	Short term results comparable with Dresden protocol. (Wen et al., 2018).
(2) Iontophoresis	(2) electric field
				More long-term comparative studies required. (Wen et al., 2018)
(3) Femto-second laser	(3) Femtosecond laser assisted
Transepithelial protocol with phonophoresis	Epithelium-on	—	Room air	An experimental procedure demonstrated a statistically significant improvement in riboflavin penetration among ultrasound treated rabbit corneas (p < 0.001) (Lamy et al., 2013).
				However, hyperthermia is a potential safety concern of this technique.
Oral riboflavin	Epithelium-on	Exposure to 15 min of direct sunlight everyday	Room air	Series of three cases of keratoconus treated in this manner described no adverse effects and corneal flattening was reported within 6 months of treatment (Jarstad et al., 2019). A small prospective study is underway.
				Limited data is available regarding dose-response relationships of systemically absorbed riboflavin and its ocular bioavailability.
Improving oxygen diffusion
Pulsed Ultraviolet-A	Epithelium-off/Variable	30 mW/cm2 for 4 min with a 1.5 s on/off cycle/Variable	Room Air	Stromal demarcation line was significantly deeper in the pulsed UVA group (213 ± 47.38 μm) compared to the continuous UVA group (149.32 ± 36.03 μm) (Moramarco et al., 2015).
				At 6 and 12 months, there was modest corneal flattening with keratometric stabilisation in 98.3% of eyes. No changes in central keratometry were noted. Moreover, mean corrected distance visual acuity, manifest refraction and endothelial cell density did not change (Gore et al., 2021)
Enhanced-fluence pulsed-light iontophoresis cross-linking	Epithelium-on with iontophoresis	18 mW/cm2 of pulsed-light on-off exposure	Room air	At 3 years, the average uncorrected distance visual acuity improved and average maximum keratometry readings decreased. Additionally, anterior segment optical coherence tomography showed that the demarcation lines were situated at an average depth of 285.8 ± 20.2 µm in more than 80% of patients at 1 month postoperatively, a value that is close to that of one created by standard epithelium-off cross-linking (Mazzotta et al., 2020b).
Periocular oxygen supplementation	Epithelium-on 0.25% riboflavin	10 J/cm2 (1 s: 1 s, pulsed)	Hyperoxic	Aydın et al. demonstrated that patients treated with periocular oxygen supplemented accelerated epithelium-on protocol experienced a larger decrease in maximum keratometry values (p-value = 0.019) and had a significantly deeper demarcation line (320 ± 17 µm) when compared to the control group (269 ± 19 µm) (Aydın and Aslan, 2021). Additionally, post-procedural endothelial cell density was comparable between both groups.
Optimising visual acuity
Customised protocol (CurV)	Epithelium-off/Epithelium-on with oxygen supplementation	Customised according to corneal topography	Room air/Hyperoxic	One year results show stronger cornea flattening and faster healing time. (Seiler et al., 2016)
				Mazzotta et al. demonstrated that high-irradiance epithelium-on customised CXL with supplemental oxygen induces visual improvement and flattens steep keratometry. Additionally, demarcation lines were approximately 30% deeper in this series than previously reported by studies which utilised epithelium-off CurV protocols. This suggests the possibility of conducting CurV without the need for de-epithelisation (Mazzotta et al., 2020b).
				Long-term studies are warranted.
Athen’s protocol	Topographically-guided transepithelial photorefractive keratectomy (PRK) followed by corneal cross-linking	6 mW/cm2; 10 min	Room air	At 3 years Athen’s protocol offered superior uncorrected distance visual acuity and flatter steep and flat keratometry than the standard epithelium-off corneal cross-linking protocol. (Kymionis et al., 2009)
Cretan protocol	Transepithelial phototherapeutic keratectomy (tPTK) with corneal cross-linking	3 mW/cm2; 30 min	Room air	A 3 years prospective comparative study of 30 eyes demonstrated vision improvement and mean reduction in corneal astigmatism. In comparison, patients who underwent the standard epithelium-off protocol did not have any improvements in visual acuity or corneal astigmatism (Grentzelos et al., 2019).
Intrastromal corneal ring segment implantation with corneal cross-linking	Intrastromal corneal ring segment implantation with corneal cross-linking	9 mW/cm2; 10 min	Room air	A large prospective study of 542 eyes showed improvements in vision and maximum keratometry value in the CXL-ICRS group (Singal et al., 2020).
Anti-infective application
PACK-CXL	Epithelium-off	3 mW/cm2; 30 min	Room air	Offered superior efficacy and healing duration in treating bacterial keratitis compared to antibiotics alone. (Tawfeek et al., 2020)
				Has a higher rate of corneal and worse visual acuity as compared to anti-fungals when treating fungal keratitis. Longer term studies are warranted. (Uddaraju et al., 2015; Prajna et al., 2020)
Thin corneas
Hypoosmolar riboflavin	Dextran-free riboflavin solution	3 mW/cm2; 30 min	Room air	Stabilised keratectasia with no resulting endothelial cell loss (Hafezi et al., 2009; Raiskup and Spoerl, 2011)
Contact-lens assisted corneal cross-linking	Iso-osmolar riboflavin 0.1% Epithelium-off	3 mW/cm2; 30 min	Room air	Achieved a stromal demarcation line with mean depth of 252.9 ± 40.8 μm. No significant endothelial loss secondary to UVA toxicity was identified (Jacob et al., 2014).
				However, the presence of a contact lens over the epithelium creates an artificial barrier that reduces oxygen diffusion into the stroma (Kling et al., 2017; Wollensak et al., 2019).
Epithelial island cross-linking technique	Customised pachymetry guided epithelial debridement	3 mW/cm2; 30 min	Room air	Technique performed on 19 eyes with improvement in vision, flattest keratometry and steepest keratometry values reported 1 year postoperatively. However, there was significant endothelial cell density loss (2550 ± 324 vs 2030 ± 200 cells/mm2) 1 year postoperatively (Cagil et al., 2017).
Epi-off-lenticule-on corneal cross-linking	Epithelium-off	3 mW/cm2; 30 min	Room air	A recent study of this technique showed that visual acuity and endothelial cell density remained stable over a 12 months follow-up. All patients were observed to have a demarcation line by 6 months follow-up (Cagini et al., 2020).
Pachymetry-based accelerated cross-linking	Various based on the nomogram	Various based on the nomogram	Various	The M nomogram was validated against clinical findings of 20 eyes (Mazzotta et al., 2018).
				However, this protocol has a distinct limitation - it requires surgeons to have access to various riboflavin formulations, and cross-linking devices that can output UVA energy at 3, 9, 15, and even 30 mW/cm2, using either continuous light or pulsed light protocols. Moreover, iontophoresis may even be required in some cases to perform the treatment.
Sub400 protocol	Epithelium-off	UV illumination time and irradiance adjusted according to the corneal thickness to achieve a safe depth of cross-linking 70 µm away from the endothelium	Room air	Pilot study showed that 90% of 39 thin corneas ranging from 214 to 398 µm achieved topographical stability at 12 months, and no eyes experienced endothelial decompensation as a result of UV irradiation toxicity. (Hafezi et al., 2020)
Other indications
LASIK Xtra and SMILE Xtra	Concomitantly with LASIK/SMILE	3 mW/cm2; 30 min	Room air	Variable short-term results reported. Longer termed and larger scale studies required. (Kanellopoulos and Asimellis, 2015; Konstantopoulos et al., 2019; Kohnen et al., 2020)
Hyper-osmolar riboflavin	Preoperative 40% glucose or intraoperative 70% glycerol	3 mW/cm2; 30 min	Room air	Reduction of central corneal thickness and visual acuity observed in patients with bullous keratopathy. (Wollensak et al., 2009; Hafezi et al., 2010)