Invasive Pharmacology Outcomes with Different Corneal Cross-Linking Protocols: A Review

Abstract

Collagen corneal cross-linking (CXL) is an invasive pharmacological treatment strategy used for corneal ectatic disorders and is currently the only treatment capable of halting the progression of the disease. In the past 20 years, the conservative management of progressive corneal ectasia has changed, thanks to this procedure that produces strengthening of the corneal tissue through the photochemical reaction generated by the combined action of riboflavin and ultraviolet A radiation. Many modified protocols have been implemented to halt the progression of the disease and to delay or prevent visual loss and surgical procedures such as corneal transplantation. Because of the variety of different protocols that are currently used, the results that are being reported are very variable, and could generate some confusion in relation to the true efficacy of the procedure. The aim of this review was to provide an overview of the aforementioned protocols that are designed to maintain the efficacy of CXL in halting the progression of the disease but avoiding the major limitations of the procedure.

Introduction

The definition of a corneal ectasia includes a process of progressive corneal thinning with alterations of the stromal collagen matrix that leads to irregular protrusion of the cornea and visual loss. Keratoconus, keratoglobus, and marginal pellucid degeneration are primary forms, whereas secondary ones are caused by refractive surgery.¹ Keratoconus is by far the most frequent and its epidemiology differs in different countries with a reported incidence and prevalence of 1.3–22.3 and 0.4–86 cases per 100,000, respectively.² In general, it is bilateral and asymmetric starting in the second decade of life with a variable rate of progression that continues until the fourth decade, when usually it becomes stable.³

Collagen corneal cross-linking (CXL) is an invasive pharmacological treatment strategy used for corneal ectatic disorders and is currently the only procedure able to halt the progression of these diseases and its primary aim is to produce a strengthening of the corneal stroma, using riboflavin combined with ultraviolet A (UVA) irradiation. The photosensitizing action of riboflavin combined with UVA irradiation (365 nm) creates additional chemical bonds between the stromal collagen fibers that increases stromal stiffness through a not fully elucidated molecular process.⁴ Standard Dresden protocol was first described by Wollensak et al. and included epithelial removal, application of 0.1% riboflavin solution for 30 min followed by 30 min of UVA irradiation.⁵ Its efficacy was demonstrated by some randomized controlled trials (RCTs).^6,7

The long period of corneal exposure and epithelial debridement are considered the cause of the most frequent complications after the procedure such as intraoperative and postoperative pain, infectious keratitis, delayed visual recovery, corneal dehydration, and abnormal wound-healing response, which are currently seen as drawbacks of the treatment.^8–10 Theoretically, all these threats can be reduced by shortening the period of corneal exposure and avoiding epithelial removal.¹¹

The corneal epithelial tight junctions limit the diffusion of riboflavin into the stroma, but its penetration can be increased using different strategies such as changing the chemical properties of the riboflavin molecule by adding chemical enhancers in its formulation^12–15 or performing a mechanical disruption of the epithelium.¹⁶ A recent strategy is iontophoresis, a noninvasive procedure that uses a small electric current to improve the penetration of charged molecules into the tissues, preserving the epithelial integrity thus achieving enough concentration of riboflavin in the stroma for the treatment.¹⁷ The complication rate in eyes treated with transepithelial CXL (TECXL) was low, as was its efficacy.^18–20

To reduce the time required in the standard procedure, the accelerated CXL (ACXL) was introduced and is based on the Bunsen–Roscoe law of photochemical reciprocity.²¹ Several recent studies showed the procedure to be safe and in most of cases effective for halting ectasia progression, despite their different protocols.^22–29 Accelerated and transepithelial protocols tend to have less postoperative pain, less corneal dehydration, and faster visual acuity (VA) recovery compared with the standard protocol and its efficacy seems to be the same.⁹ However, apparently the failure rate tends to be greater with these modified protocols.³⁰

We believe that corneal collagen cross-linking is an invasive pharmacological therapy and because of the variety of different protocols currently used, numerous results exist, which only cause confusion regarding the outcomes. Therefore, the aim of this review was to provide an overview of the aforementioned protocols that are designed to maintain the efficacy of CXL in halting the progression of the disease but avoiding the major limitations of the procedure.^8,9

Methods

A computer-assisted search of English language articles was performed using the PubMed index. Varying combinations of the terms “accelerated CXL,” “trans-epithelial CXL” “conventional, epithelium-off, epithelium-without CXL,” “high-tense, high-fluence CXL” and “epithelium-on, epithelium-with CXL” “iontophoresis assisted CXL” were used and cross-checked with references from related articles. Our searching data ranged from January 2012 to August 2017. All studies were considered including retrospective case series (RCS), prospective case series (PCS), or RCTs. No articles were omitted. We did not include “gray literature” in this review that is defined as literature that is not available through the usual bibliographic sources such as databases or indexes.

TECXL was defined as intact corneal epithelium with or without methods used to change epithelial barrier permeability; ACXL was defined as intensity >3 mW/cm² and exposure duration <30 min irrespective of the specific parameters used in each study.

Study characteristics were extracted by a single reviewer and tabulated such as study design, participant demographics, definition of progressive keratoconus, details of the intervention (eg, riboflavin ingredients and frequency of riboflavin instilment, wavelength of UVA, intensity, and radiation duration), clinical outcomes and adverse events.

Because the aim of the treatment is to halt the progression of the disease, maximum keratometry (Kmax) variation after the procedure was used as a primary outcome. Best-corrected visual acuity (BCVA), depth of demarcation line (DDL), failure rate (percentage of eyes that continued to progress), and loss of vision were used as secondary outcomes.

Kmax variation, VA (logMAR), DDL, failure rate, and loss of vision were all continuous data. Data were tabulated and letters were assigned for the results such as a significant improvement (I) or reduction (R), nonsignificant improvement (i) or reduction (r), worsening (W) and stabilization (U) for Kmax and BCVA. A single reviewer performed the scientific literature search. Primary selection was conducted through browsing titles and abstracts, and then the full copies of the remaining studies were obtained to determine what data would be included in this review article.

We used the Scottish Intercollegiate Guideline Network (SIGN) for grading the level of evidence for all the studies and this is displayed in all tables, next to the study type.

Protocols

Conventional corneal cross-linking

The Dresden protocol initially described by Wollensak et al.⁵ included primarily epithelial removal, the application of 0.1% riboflavin solution for 30 min followed by 30 min of UVA irradiation with a wavelength of 370 nm, and a power of 3 mW/cm² (5.4 J/cm²)

Accelerated corneal cross-linking

This protocol uses a fraction of the irradiation time of the original protocol. To achieve the total energy levels equal to those proposed by the standard procedure, the irradiance intensity is increased thus reducing the time of exposure to 10, 5, or 3 min. Accordingly, irradiance is 9 mW/cm² for the 10-min treatment, 18 mW/cm² for the 5-min treatment, 30 mW/cm² for the 3-min treatment, but variability exists depending on the author. This technique is based on the Bunsen–Roscoe law of photochemical reciprocity.²¹

Transepithelial (Epi on) corneal cross-linking

This procedure does not include epithelial removal. Riboflavin is a hydrophilic molecule and as such is unable to pass through the tight junctions of an intact epithelium. Enhancer agents used to increase the penetration of riboflavin through the stroma such as trometamol (tris), EDTA (ethylenediaminetetraacetic acid), BAK (benzalkonium chloride), proparacaine, sodium chloride, and even vitamin E³¹ are utilized in this protocol before irradiation.

Iontophoretic transepithelial corneal cross-linking

This technique facilitates drug penetration into tissue using a small electric current. Riboflavin is an appropriate molecule for iontophoretic transfer as it is small, negatively charged at physiologic pH and easily soluble in water.¹⁷ The device uses a constant current generator of 1 mA for 5 min (total dose of 5 mA/5 min) and then radiation is applied.

Pulsed corneal cross-linking

Oxygen plays a fundamental role in CXL reaction^32,33: Oxygen concentration in the cornea is modulated by UVA irradiance and rapidly decreased at the beginning of the exposure. Some authors suggest that high oxygen availability potentially increases the overall efficacy of the procedure. Protocols with accelerated methods deliver UV light with an on-off pattern that would enable a better diffusion of the oxygen into the corneal stroma and possibly a greater effect.³⁴

Rivoflavin

Riboflavin creates a coating on the ocular surface during the initial phase of the procedure and acts as a shield by absorbing great amount of UVA light, reducing its total dose, and also the depth of the CXL effect in the corneal stroma.⁴

Iso-osmolar riboflavin (0.1% riboflavin 5 phosphate with 20% of dextran T-500) was used in the conventional CXL protocol; however, the oncotic effect of Dextran, a 500 kDa polyglucose biopolymer with a high affinity to water, resulted in some corneal deswelling and thinning.⁵

Dextran-free riboflavin was introduced avoiding the oncotic effects of the biopolymer; dextran-free and hydroxypropyl methylcellulose (HPMC) solutions showed less corneal thinning.³⁵

Hypoosmolar riboflavin is used in thin corneas by swelling the corneas to increase stromal thickness before UV irradiation; because the hydrophilic property of the deepithelialized corneal stromal proteoglycan can swell to twice its normal size when it is irrigated with a hypoosmolar solution.^36,37

The wide variety of different instillation techniques and composition of the solution that are currently used for this procedure are described in Tables 1–8.

Table 1.

Noncomparative Studies from Accelerated Cross-Linking Protocols (Evidence Level 3, Grade of Recommendation D)

First author (year) Ref no.	Technique	Riboflavin time of soaking and application during irradiation	Intensity (mW/cm²) time/area	Type of study	FU (months)	No. of eyes	BCVA	Kmax	DDL	Progression (%)	Loss of vision (%)	Indication
Cınar (2014)²⁹	ACXL	Iso 0.1% HPMC; every 3 × 30 min and every 2 min	9 mW/cm² × 10 min/8 mm	PCS (3)	6	23	I	R	NR	NR	17	KC
Shetty (2014)²⁷	ACXL	Iso 0.1% Dx 20%; every 2 × 30 min and every 2 min	9 mW/cm² × 10 min/8 mm	PCS (3)	24	18	I	R	NR	17	NR	KC Ped
Elbaz (2014)²⁸	ACXL	Iso 0.1% Dx 20%/every 2 × 30 min and every 2 min	9 mW/cm² × 10 min/8 mm	RCS (3)	12	16	U	U	NR	12	0	KC
Ozgurhan (2015)²³	ACXL	Iso 0.1% Dx 20%/every 2 × 15 min	30 mW/cm² × 3 min/8 mm	RCS (3)	12	34	i	R	205	0	0	KC thin
Chan (2015)²²	ACXL	Iso 0.1% Dx 20%/every 2 × 30 min and every 2 min	18 mW/cm² × 5 min/9 mm	PCS (3)	12	25	U	R	NR	Doubt	NR	KC
Waszczykowska (2015)²⁵	ACXL	Iso 0.1% Dx 20%/every 2 × 30 min and every 3 min	6 mW/cm² × 15 min/8 mm	PCS (3)	24	16	U	U	282	6	6	KC
Alnawaiseh (2015)⁶⁹	ACXL	Iso 0.1%/every 2 × 30 min and every 2 min	18 mW/cm² × 5 min	RCS (3)	21	28	U	R	NR	NR	NR	KC
Badawi (2017)⁴⁵	ACXL	Iso 0.1% HPMC/every 2 × 10 min and every 2 min	10 mW/cm² × 9 min/8 mm	PCS (3)	12	33	I	R	NR	0	NR	KC Ped

Level of evidence (next to study type).

ACXL, accelerated corneal cross-linking; BCVA, best-corrected visual acuity; CXL, corneal cross-linking (Standard procedure); DDL, depth of demarcation line; Dx, dextran; FU, follow-up; HPMC, hydroxypropyl methylcellulose; I, significant improvement; i, nonsignificant improvement; Iso, isotonic riboflavin; KC, progressive keratoconus; KC thin, thin corneas; NR, no reported; PCS, prospective case series; Ped, pediatric cases; R, significant reduction; r, nonsignificant reduction; RCS, retrospective case series; U, unchanged; W, worsening.

Table 2.

Comparative Studies from Accelerated Protocols (Evidence Level 2−, Grade of Recommendation D)

First author (year)	Technique	Riboflavin before and during treatment	Intensity (mW/cm²) × duration	Type of study	FU (months)	Eyes	BCVA	Kmax	DDL	Failure	Loss of vision	Indication
Kanellopoulos (2012)⁷⁰	CXL vs. ACXL contralateral	Iso 0.1%; every 0.5 × 5 min	7 mW/cm² × 15 min	PCR (3)	18	21	I	R	NR	0%	NR	KC
Razmjoo (2017)⁵¹	CXL vs. ACXL			PCI (2++)	6	40	NIGD	NIGD	NR	NR	NR	KC
		NR	3 mW/cm² × 30 min			20	U	U
		NR	18 mW/cm² × 5 min			20	i	U
Mazzotta (2014)³⁵	ACXL vs. pulsed ACXL	Iso 0.1% Dx free + HPMC × 10 min		PCI (2−)	12	20				NR	NR	KC
			30 mW/cm² × 8 min/pulse			10	i	R	217
			30 mW/cm² × 4 min/continuous			10	i	r	182
Brittingham (2014)⁴⁴	CXL vs. ACXL			RCS (2−)	12	131		IGD	NR		NR	KC
		Iso 0.1% Dx 20% or Hypo 0.1%; every 5 × 20 min and every 2 min	3 mW/cm² × 30 min			81		r	323
		Iso 0.1% Dx 20% or Hypo 0.1%; every 5 × 20 min and every 2 min	9 mW/cm² × 10 min			50		W	245
Tomita (2014)⁷¹	CXL vs. ACXL			PC (2−)	12	48	NIGD	NIGD	NR		NR	KC
		Iso 0.1% Dx 20% × 30 min	3 mW/cm² × 30 min			18	i	R	380
		Iso 0.1% HPMC × 15 min	30 mW/cm² × 3 min			30	i	r	294
Cinar (2014)⁵⁷	CXL vs. ACXL			PCI (2−)	6	26	NIGD	NIGD	NR	NR	20%	KC
		Iso 0.1% Dx 20%; every 3 × 30 min and every 5 min	3 mW/cm² × 30 min			13	i	R			15
		Iso 0.1% HPMC; every 3 × 30 min and every 2 min	9 mW/cm² × 10 min			13	I	R			21
Chow (2015)⁷²	CXL vs. ACXL	Iso 0.1% Dx 20%; every 2 × 30 min and every 2 min		PCI (2−)	12	38	NIGD	NIGD	NR	NR	NR	KC
			3 mW/cm² × 30 min			19	I	R
			18 mW/cm² × 5 min			19	I	r
Moramarco (2015)⁴⁰	ACXL vs. pulsed ACXL	Iso 0.1% HPMC 1%; every 1–2 × 10 min		RCS (2−)	6	60	No	No		NR	NR	KC
			30 mW/cm² × 8 min/pulse			30			213
			30 mW/cm² × 4 min/continuous			30			149
Ng (2016)⁶¹	CXL vs. ACXL			RCS (2−)	12	26	NIGD	IGD			NR	KC
		Iso 0.1% Dx 20%; every 2 × 25 and every 5 min	3 mW/cm² × 30 min			14	I 0.13	R	282	0%
		Iso 0.1% Dx 20%; every 2 × 25 and every 5 min	9 mW/cm² × 10 min			12	U	U	209	8%

Level of evidence (next to study type).

Hypo, hypotonic riboflavin; IGD, intergroup difference; NIGD, no intergroup difference; PCI, prospective comparative interventional; PCR, prospective comparative randomized.

Table 3.

Comparative Studies from Accelerated Protocols (Evidence Level 2+, Grade of Recommendation C)

First author (year)^{Ref. No.}	Technique	Riboflavin before and during treatment	Intensity (mW/cm²) × duration	Type of study	FU (months)	Eyes	BCVA	Kmax	DDL	Failure	Loss of vision	Indication
Sadoughi (2018)⁶⁰	CXL vs. ACXL contralateral	Iso 0.1% Dx 20%; every 2 × 30 and every 5 min		PCI (2+)	18	30	NIGD	NIGD	NR	NR	NR	KC
			3 mW/cm² × 30 min			15	i	r
			9 mW/cm² × 10 min			15	i	r
Cummings (2016)⁶⁶	CXL vs. ACXL	Iso 0.1% Dx 20%; every 1 × 20–30 min		RCI (2−)	12	100			NR	NR		KC
			3 mW/cm² × 30 min			66	I	R			1.5%
			9 mW/cm² × 10 min			34	I	r
Yıldırım (2017)⁶⁷	ACXL vs. ACXL			PCI (2++)	12		NIGD	NIGD	NR	4%	4%	KC
		Iso 0.1% Dx 20%; every 2 × 20 min	18 mW/cm² × 5 min			74	i	R
		0.1% Dx 20%; every 2 × 20 min	30 mW/cm² × 3 min			72	i	R
Woo (2017)⁷³	CXL vs. ACXL			PCI (2+)	12	76	NIGD	NIGD	NR	NR	NR	KC
		Iso 0.1% Dx 20%; every 2 × 30 min and every 2 min	3 mW/cm² × 30 min.			29	I	r
		Iso 0.1% Dx free; every 2 × 10 min	30 mW/cm² × 4 min			47	I	U
Jiang (2017)³⁴	CXL vs. pulsed ACXL			PCI (2+)	12	72	NIGD	NIGD				KC
		Iso 0.1% Dx 20%; every 3 × 30 min and every 3 min	3 mW/cm² × 30 min			36	I	R	284	6%
		Iso 0.1% Dx free + HPMC; every 2 × 10 min	30 mW/cm² × 8 min/pulse			36	I	R	201	11%
Baenninger (2017)³⁰	CXL vs. ACXL			RCS (2−)	12	78	NIGD	NIGD	NR	19%	1.30%	KC Ped
		Iso 0.1% Dx 20%; every 3 × 30 min and every 5 min	3 mW/cm² × 30 min			39	I	r		23%
		Iso 0.1% Dx 20%; every 3 × 30 min and every 5 min	9 mW/cm² × 10 min			39	I	r		15%

Level of evidence (next to study type).

Table 4.

Randomized Controlled Trials from Accelerated Cross-Linking Protocols (Evidence Level 1− to 1+, Grade of Recommendation B)

First author (year)^{Ref. No.}	Technique	Riboflavin before and during treatment	Intensity (mW/cm²) × duration	Type of study	FU (months)	Eyes	BCVA	Kmax	DDL	Failure	Loss of vision	Indication
Sherif (2014)⁵⁸	CXL vs. ACXL			RCT (1−)	12	25		NIGD	NR	NR	NR	KC
		Iso 0.1% Dx 20%; Every 2 × 30 and every 2 min	3 mW/cm² × 30 min			11	I	R
		Iso 0.1% Dx 20%; every 2 × 30 min and not during irradiation	30 mW/cm² × 4 min			14	I	R
Hashemian (2014)³⁸	CXL vs. ACXL			RCT (1−)	15	153	NIGD	NIGD	NR	NR	NR	KC
		Iso 0.1% Dx 20% × 20 min	3 mW/cm² × 30 min			76	I	R
		Iso 0.1% Dx 20% × 20 min	30 mW/cm² × 3 min			77	I	R
Shetty (2015)⁵⁹	CXL vs. ACXL	Iso 0.1% Dx 20%; every 2 × 30 and every 2 min		RCT (2++)	12	138	IGD	IGD		1.4%	NR	KC
			3 mW/cm² × 30 min			36	I 0.04	R 1.32	280
			9 mW/cm² × 10 min			36	I 0.06	R 0.67	292
			18 mW/cm² × 5 min			33	I 0.10	R 0.52	203	3%
			30 mW/cm² × 3 min			33	i 0.02	r 0.18	201	3%
Hashemi (2015)⁵²	CXL vs. ACXL contralateral	Iso 0.1% Dx 20%; every 3 × 30 min and every 3 min		RCT single mask (1+)	18	62	NIGD	NIGD	NR	NR	NR	KC
			3 mW/cm² × 30 min			31	U	R
			18 mW/cm² × 5 min			31	U	U

Level of evidence (next to study type).

RCT, randomized controlled trial.

Table 5.

Noncomparative Studies from Transepithelial Protocols (Evidence Level 3, Grade of Recommendation D)

First author (year)^{Ref. No.}	Technique	Riboflavin before and during treatment	Intensity (mW/cm²) × duration	Type of study	FU (months)	Eyes	BCVA	Kmax	DDL	Failure	Loss of vision	Indication
Koppen (2012)⁴⁹	TECXL	Propa 0.5% + BAC 0.005% every 3 × 30 then Iso 0.1% Dx 20% every 3 × 30 and during 3 × 30 min	3 mW/cm² × 30 min	PCS	6	53	I	W	NR	NR	NR	KC
Spadea (2012)⁵⁰	TECXL	Iso 0.1% Dx 15% + EDTA/tris every 10 × 60 then every 3 × 15 and every 5 min during irradiation	3 mW/cm² × 30 min	PCS	6	16	i	R	NR	NR	NR	KC
Filippello (2012)⁷⁴	TECXL	Iso 0.1%, Dx 20% + tris + EDTA; every 10 × 30 and every 5 min	3 mW/cm² × 30 min	PCS	18	20	I	R	100	No	No	KC
Caporossi (2013)¹³	TECXL	Iso 0.1%, Dx 15% + tris + EDTA; every 2 × 30 and every 5 min	3 mW/cm² × 30 min	PCS	24	26	U	W	NR	19%	19%	KC
Salman (2013)⁷⁵	TECXL vs. CXL (contralateral)	Iso 0.1%, Dx + tris + EDTA every 2 × 30 and every 5 min	3 mW/cm² × 30 min/9 mm	PCI (2−)	12	22	i	r	NR	NR	4.00%	KC Ped
Lesniak (2014)⁴⁸	TECXL	BAK 0.01% every 5 × 30 then Iso 0.1% Dx free every 2 × 30 and every 2 min during irradiation	3 mW/cm² × 30 min	PCS	6	30	I	R	NR	3%	3%	KC
De Bernardo (2014)⁴⁷	TECXL	Iso 0.1% Dx 15% + EDTA tris; every 5 × 30 and every 3 min	3 mW/cm² × 30 min	PCS	6	30	I	U	NR	NR	NR	KC
Khairy (2014)⁷⁶	TECXL	Hypo 0.1% × 30 min	3 mW/cm² × 30 min	PCS	12	32	U	R	NR	NR	NR	KC
Caruso (2016)³¹	TECXL	Iso 0.1% Dx 20% + Vit E/pool × 15 min and not during irradiation	1.5 mW/cm² × 10 min (customized power)	PCS	24	25	I	R	NR	20%	NR	KC
Zhang (2016)⁴¹	ATECXL	Ribo-5-phosphate 0.25%, HPMC, BAC + NaCl + tris Pool × 10 min	45 mW/cm² × 5 min 1	PCS	12	28	i	r	NR	NR	NR	KC
Artola (2017)⁴²	ATECXL pulse	Hypo Dx free + BAK; every 1 × 4 then Ribo 25% every 1 × 6 min	45 mW/cm² × 2.5 min pulse (2 sec ON/1 sec OFF)	PCS	12	19	U	r	205	NR	NR	KC
Akbar (2017)⁷⁷	ATECXL	0.25% BAK + EDTA + tris + HPMC every 1 × 4 and Ribo 25% every 1 × 6 min	9 mW/cm² × 10 min	PCS	12	26	U	R	No	10%	NR	KC
Aixinjueluo (2017)⁷⁸	TECXL	Ribo 0.25 + BAK + EDTA + tris + HPMC; every 1 × 4 then Ribo 25% every 1 × 6 min and every 1 min	30 mW/cm² × 3 min	PCS	12	30	I	R	170	NR	NR	KC

Level of evidence (next to study type).

ATECXL, accelerated transepithelial corneal cross-linking; BAK, benzalkonium chloride; EDTA, ethylenediaminetetraacetic acid; TECXL, transepithelial corneal cross-linking; tris, trometamol.

Table 6.

Comparative Studies from Transepithelial Protocols (Level of Evidence 2− and 2+, Grade of Recommendation C)

First author (year)^{Ref. No.}	Technique	Riboflavin before and during treatment	Intensity (mW/cm²) × duration/area	Type of study	FU (months)	Eyes	BCVA	Kmax	DDL	Failure	Loss of vision	Indication
Magli (2013)⁶⁵	CXL vs. TECXL			RCS (2−)	12	39	NIGD	NIGD	NR	NR	NR	KC Ped
		Iso 0.1% Dx 20%; every 2 × 30 and every 2 min	3 mW/cm² × 30 min/9 mm			23	U	R
		Iso 0.1%, Dx 20% + tris + EDTA; every 5 × 30 and every 5 min	3 mW/cm² × 30 min			16	U	R
Kocak (2014)⁶²	CXL vs. TECXL			RCS (2−)	20	36			NR		NR	KC
		Iso 0.1% Dx 20%; every 2 × 30 and every 2 min	3 mW/cm² × 30 min/9 mm			19	U	R		11%
		Iso 0.1% Dx 15% EDTA + HPMC; every 2 × 30 and every 2 min	3 mW/cm² × 30 min			17	U	W		65%
Rossi (2015)⁵³	CXL vs. TECXL			PCI (2+)	12	20	IGD	NIGD	NR	NR	NR	KC
		Iso 0.1% Dx 20%; every 2 × 30 and every 2 min	3 mW/cm² × 30 min/9 mm			10	I	R
		Iso 0.1%, Dx 15% + tris + EDTA; every 5 × 30 and every 5 min	3 mW/cm² × 30 min			10	W	R
Çerman (2015)⁶⁸	CXL vs. TECXL			RCS (2−)	18							KC
		Iso 0.1% Dx 20%; every 3 × 30 and every 2 min	3 mW/cm² × 30 min/9 mm				I	R	267	6%	0
		Iso 0.1%, Dx 15% + tris + EDTA; every 3 × 30 and every 2 min	3 mW/cm² × 30 min				I	U	125	22%	0
Eraslan (2017)⁶³	CXL vs. TECXL			PCI (2−)	24	36						KC Ped
		Iso 0.1% Dx 20%; every 2 × 30 and every 2 min	3 mW/cm² × 30 min/9 mm			18	I	R	272	5%	12%
		Iso 0.25%–1.2% HPMC + BAK; every 2 × 30 and every 2 min	3 mW/cm² × 30 min			18	I	U	136	33.3%	5%
Henriquez (2017)⁶⁴	CXL vs. ATECXL			PC (2+)	12	61				8%	0%	KC Ped
		Iso 0.1% Dx 20%; every 5 × 30 and every 5 min	3 mW/cm² × 30 min			25	I	r		12%	0
		Iso 0.25% HPMC 1.0% BAK 0.007%; every 5 × 30 min and not during irradiation	18 mW/cm² × 5 min			36	I	U		5%	0

Level of evidence (next to study type).

Table 7.

Randomized Controlled Trials from Accelerated Cross-Linking Protocols (Level of Evidence 1− and 1+, Grade of Recommendation B)

First author (year)^{Ref. No.}	Technique	Riboflavin before and during treatment	Intensity (mW/cm²) × duration/area	Type of study	FU (months)	Eyes	BCVA	Kmax	DDL	Failure	Loss of vision	Indication
Stojanovic (2014)³⁶	CXL vs. TECXL			RCT (1−)	12	40			NR	NR	NR	KC
		Propa 0.5% BAC 0.005% + Hypo 0.5% Dx free; every 30 s × 15 and balanced solution every 3 min	3 mW/cm² × 30 min/8 mm			20	I	U
		Propa 0.5% BAC 0.005% + Hypo 0.5% Dx free; every 30 s × 15 and balanced solution every 3 min	3 mW/cm² × 30 min			20	I	U
Soeters (2015)¹⁵	CXL vs. TECXL			RCT (1+)	12	61	NIGD	IGD				KC
		Iso 0.1% Dx 20%; every 3 × 30 and every 5 min	3 mW/cm² × 30 min/9 mm			26	I	R	266	0%	0%
		Iso 0.1%, Dx 20% + EDTA; every 2 × 15 then 15 pool and every 5 min during irradiation	3 mW/cm² × 30 min			35	I	U	No visible	23%
Nawaz (2015)⁴³	CXL vs. TECXL			RCT (1−)	6	40			NR	NR	NR	KC
		Iso 0.1% Dx 20%; every 3 × 30 and every 3 min	3 mW/cm² × 30 min/7 mm			20	I	R
		Propa 0.5% BAC 0.005% + Iso 0.1% Dx 20%; every 3 × 30 and every 3 min	3 mW/cm² × 30 min			20	I	R
Al Fayez (2015)⁴⁶	CXL vs. TECXL			RCT (1−)	36	74	IGD	IGD				KC
		Iso 0.1% Dx 20%; every 3 × 30 and every 3 min	3 mW/cm² × 30 min/9 mm			36	I	R	320	0%	NR	KC
		Propa 0.5% BAC 0.02% + Iso 0.1% Dx free; every 2 × 30 and every 3 min	3 mW/cm² × 30 min			36	W	W	190	55%

Level of evidence (next to study type).

Table 8.

All Studies from Iontophoretic Transepithelial Corneal Cross-Linking (Higher Level of Evidence 1+, Higher Grade of Recommendation B)

First author (year)^{Ref. No.}	Technique	Riboflavin before and during treatment	Intensity (mW/cm²) × duration/area	Type of study	FU (months)	Eyes	BCVA	Kmax	DDL	Failure	Loss of vision	Indication
Magli (2016)³⁹	TICXL	Iso 0.1% Dx free + EDTA tris + current × 5 min	10 mW/cm² × 9 min	PCS (3)	18	13	U	U	NR	NR	NR	KC Ped
Buzzonetti (2015)⁵⁶	TICXL	Hypo 0.1% Dx free + EDTA tris + current × 5 min	10 mW/cm² × 9 min	PCS (3)	15	14	U	U	180	0%	NR	KC Ped
Bikbova (2016)²⁰	CXL vs. TICXL			RCT (1−)	24	149	IGD	IGD		NR	NR	KC
		Iso 0.1% Dx 20%; every 3 × 30 and every 2 min	3 mW/cm² × 30 min/9 mm			62	I	R − 2.15	287
		Hypo 0.1% + current × 10 min. and every 2 min during irradiation	3 mW/cm² × 30 min			57	I	R 0.97	169
Jouve (2016)⁵⁴	CXL vs. TICXL			PCI (2+)	24	80	IGD				NR	KC
		Iso 0.1% Dx 20%; every 2 × 20 and every 5 min	3 mW/cm² × 30 min/9 mm			40	I	R	291	8%
		Hypo 0.1% Dx free + EDTA tris + current × 5 min	10 mW/cm² × 9 min			40	I	U	216	20%
Lombardo (2016)⁵⁵	CXL vs. TICXL			RCT (1+)	12	34	NIGD		NR			KC
		Iso 0.1% Dx 20%/every 3 × 30 and 0.025% every 3 min	3 mW/cm² × 30 min/9 mm			12	i	R 0.5		8%	8%
		Hypo 0.1% Dx free + EDTA tris + current × 5 min	10 mW/cm² × 9 min			22	I	R 0.8		18%
Vinciguerra (2016)⁷⁹	CXL vs. TICXL			PCI (2+)	12	40			NR	NR	NR	KC
		Iso 0.1% Dx 20%; every 1 × 30 min	3 mW/cm² × 30 min/9 mm			20	I	R
		Hypo 0.1% Dx free + EDTA tris + current × 5 min	10 mW/cm² × 9 min			20	I	r
Cantemir (2017)⁸⁰	CXL vs. TICXL			RCS cohort (2−)	24	80			NR	NR	NR	KC
		Iso 0.1% Dx 20% every 3 × 15 min	10 mW/cm² × 9 min			40	I	R 1.2
		Hypo 0.1% Dx free + EDTA tris + current × 5 min	10 mW/cm² × 9 min			40	I	R 0.9

Level of evidence (next to study type).

TICXL, iontophoretic transepithelial corneal cross-linking.

Results

A total of 580 records were retrieved by searching the electronic databases and by indexing references of related literature. There were ∼31 duplications and 470 obviously unrelated records that were recognized by titles and abstracts. We excluded 20 citations by browsing full texts, and 57 eligible studies were included finally.

Among these studies, 27 (48%) corresponded to ACXL (Tables 1–4), 23 (40%) to TECXL (Tables 5–7), and 7 (12%) iontophoretic transepithelial corneal cross-linking (TICXL) (Table 8).

In relation to the study design, 10 studies (17%) were RCTs, 20 studies (36%) were noncomparative PCS, 19 (33%) were comparative prospective cases, and 10 studies (19%) were RCS. Eight studies (14%) from the ACXL protocol were noncomparative and 3 of them (5%) were retrospective; 19 studies (34%) were comparative and among them 3 (5%) were RCS and 4 (7%) were RCTs. Twelve studies (21%) (Table 6) from the TECXL protocol were noncomparative and from 11 (19%) comparative studies, 4 (7%) (Table 7) were RCTs. In the TICXL group, 4 (7%) studies were PCS, 2 (3.5%) were RCTs and 1 (1.7%) was a retrospective cohort study (Table 8).

The sample size varied widely among the studies, the largest sample size enrolled 153 patients (153 eyes),³⁸ the smallest one had 13 patients (13 eyes),³⁹ and the sample sizes for most of the studies were ∼30–70 eyes. All eligible studies included both genders and enrolled progressive keratoconus patients (which was defined as an increase in >0.75 of the Kmax within 6 months to 1 year of follow up) as the participants in most of the studies. Nine studies (15%) (Tables 1, 3, 5, and 8) evaluated pediatric cases (<18 years old), the others (85%) were all adult patients and only 1 (1.6%) study evaluated thin corneas (treated with iso-osmolar riboflavin).

All standard protocols in the studies (Table 1–8) used an irradiance of 3 mW/cm² for 30 min of UVA. However, the combinations of duration and intensity used for accelerated protocols were very different among them, for example, 30 mW/cm² for 3 min, 30 mW/cm² for 4 min, 18 mW/cm² for 5 min, 9 mW/cm² for 10 min, and so on. Only 5 studies^{34,35,40–42} used a pulsed protocol. The wavelength of 365 or 370 nm was the most commonly used except for 1 study that used 765 nm.⁴³

The soaking procedure was also very different among the studies ranging from every 2 min for 5 min to every 3 min for 30 min; during irradiation and varied from every 1–5 min, and even some without application.³¹ Almost all the ACXL protocols used isotonic dextran free with HPMC riboflavin solutions, 2 studies used hypotonic solutions^36,44 for thin corneas, and only 1 study used isotonic solution for these patients.²³ The riboflavin used in transepithelial procedure was different from that in standard and accelerated methods, containing some loosened or permeable ingredients, such as EDTA, tromethamine, and others (Tables 5–7). Excluding TECXL, all studies scraped the corneal epithelium mechanically, but some used chemical means to remove epithelium such as ethanol or topical anesthetics^28,29,45 and the diameter of epithelium removal varied in a range of 8–10 mm in all the studies.

The follow-up time varied among the studies, the longest being 36 months (Al Fayez et al.),⁴⁶ the shortest 6 months,^{29,40,43,47–51} and most of them were ∼12–24 months.

Primary and secondary evaluated outcomes

Primary outcome, Kmax variation

The Kmax reduction overall was significant in almost half of the studies (49%), from these the follow-up time ranged from 12 to 24 months in 95%. Fifteen studies (26.3%) remained stable [6, 6, and 3 for ACXL, TECXL, and TICXL, respectively) with a mean follow-up of 15 months. Nonsignificant reductions occurred in 9 (16%) studies (6, 3, and none for ACXL, TECXL, and TICXL, respectively)] with a mean follow-up of 15 months. On the contrary, only 5 studies (8.7%) reported worsening of this variable with a mean follow-up of 19 months, 1 in the ACXL and 4 in the TECXL group, in the latter 1 was an RCT.⁴⁶

Secondary outcomes: BCVA, DDL, failure rate

BCVA significantly improved in 50.8% of all studies, 11 (19%), 12 (21%), and 5 studies (8.7%) for ACXL, TECXL, and TICXL, respectively. No significant improvement occurred in 10 (16%) studies (7, 3, and none for ACXL, TECXL, and TICXL, respectively). It remained unchanged in 15 (26.2%) studies (7, 6, and 2 for each group, respectively); from these, 1 was an RCT⁵² in the ACXL group. Only 2 (3.5%) studies^46,53 showed worsening of BCVA in the TECXL group and 2 (3.5%) studies^40,44 did not show VA.

With respect to the studies that evaluated the DDL, overall only 20 (35%) reported it; for the ACXL protocols this depth ranged from 201 to 294 μm in comparison with the conventional one that ranged from 284 to 323 μm. In the TECXL and TICXL groups, 9 (30%) studies report it, ranging from 100 to 205 μm and 1 study¹⁵ did not find a visible line at 1 and 3 months.

With respect to complications after the procedure, overall 24 (42%) studies reported progression or failure (defined as an increase of 1 diopter in the Kmax during the follow-up) and 16 (28%) studies reported vision loss (defined as loss of more than 1 line during the follow-up). Failure rate was reported in 11, 10, and 3 studies for ACXL, TECXL, and TICXL protocols, respectively; these results ranged from 0% to 17%²⁷ in the ACXL group, 5% to 55%⁴⁶ in the TECXL, and 0% to 20%⁵⁴ in the TICXL.^54–56 On the contrary, vision loss ranged from 0% to 21%.^29,57 in 8 studies for ACXL, 0% to 19%¹³ in 7 studies corresponding to TECXL, and 0% in the TICXL group.

Discussion

This investigation was performed with the main purpose of analyzing the different collagen cross-linking protocols that are used nowadays for the treatment of corneal ectatic disorders. From the many studies that are published in the scientific literature, it was found that a significant number of different variables are included and analyzed in all of them. Thus, conducting a proper comparison among the aforementioned investigations will clearly have a methodological bias. Therefore, the main purpose of the current scientific review was to provide a general perspective in the state of the art of the different approaches in corneal collagen cross-linking in the treatment of corneal ectatic disorders.

Primary outcome

Change in maximum keratometry (Kmax variation)

Accelerated corneal cross-linking

We used Kmax at the last follow-up as our primary outcome because it is a widely used parameter that demonstrates progression of the ectasia. From the 4 RCTs, Hashemian et al.³⁸ (153 eyes—15 months) used the shortest exposure time (3 min with 30 mW/cm² of fluence) achieving a stabilization of the disease during the follow-up with a significant reduction in Kmax in both groups with no intergroup difference. Sherif⁵⁸ (25 eyes—12 months) used the same fluence for 4 min and found a significant decrease in the Kmax in both groups with no significant difference between them during the follow-up time, but the sample was small.

Shetty et al.⁵⁹ (138 eyes—12 months) compared 3 different fluences with the conventional procedure, and reported a significant reduction in the Kmax in all groups, except for the group with the highest fluence and shortest time (30 mW/cm² × 3 min); it is important to observe that the flattening effect was significantly greater in the conventional group compared with the others.

Hashemi et al.⁵² (62 eyes—18 months) compared 18 mW for 5-min protocol with the conventional group and the Kmax remained unchanged during the follow-up, but there was a significant flattening in the conventional group.

On the contrary, most of the prospective comparative studies showed stabilization or a nonsignificant decrease in Kmax (Tables 2 and 3) during the follow-up. From those with the longest follow-up time, Sadoughi et al.⁶⁰ (30 eyes—18 months) reported a nonsignificant decrease in Kmax in both groups. Ng et al.⁶¹ (26 eyes—14 months) showed an unchanged Kmax but the conventional group had a significant reduction; both authors used a 9 mW/cm² × 10-min protocol. Baenninger et al.³⁰ (78 eyes—12 months) retrospectively reviewed pediatric patients and found stabilization with a nonsignificant reduction in Kmax for both groups. Badawi⁴⁵ and Shetty et al.²⁷ also evaluated pediatric cases for 12 and 24 months, respectively, reporting a significant flattening during the follow-up.

From the group that performed pulsed accelerated protocols; Jiang et al.³⁴ (72 eyes—12 months) used a 1 s on-/1 s off-pulsed protocol and compared it with the conventional one reporting a significant reduction in Kmax for both groups with no differences between them. Mazzotta et al.³⁵ (20 eyes—12 months) used a 1 s on-/1 s off-pulsed protocol and compared it with a continuous accelerated 30 mW/cm² × 4-min protocol finding a significant flattening in the pulsed group compared with the conventional one.

Transepithelial corneal cross-linking

We found 4 RCTs (Table 7) that compared it with the standard protocol, Stojanovic et al.³⁶ (40 eyes—12 months) used a hypotonic riboflavin in both groups and did not find a significant decrease in Kmax in either of them. Soeters et al.¹⁵ (61 eyes—12 months) found stabilization of the Kmax in the Epi on group but the conventional group showed a significant reduction. Nawaz et al.⁴³ (40 eyes—6 months) achieved a significant decrease in Kmax in both groups. Caporossi et al.,¹³ Kocak et al.,⁶² and Al Fayez et al.⁴⁶ (74 eyes) showed a worsening of Kmax at 24, 12, and 36 months of follow-up, respectively.

Four studies evaluated pediatric cases; Salman (22 eyes) found no significant Kmax reduction. Magli et al. (39 eyes) retrospectively compared it with the standard protocol and found a significant reduction in both groups. Eraslan et al. (24 eyes) prospectively found no change in comparison with the standard protocol, the latter had a significant reduction. Henriquez et al. (61 eyes) used an accelerated epi on protocol, with stabilization in both groups throughout the 12-month follow-up. Several other studies showed stabilization or a nonsignificant decrease in Kmax (Table 6).

Iontophoretic transepithelial corneal cross-linking

Two RCTs were found; Bikbova and Bikbov²⁰ (149 eyes—24 months) reported a significant Kmax reduction in both groups but it was significantly greater in the standard protocol. Lombardo et al.⁵⁵ (34 eyes—12 months) achieved a significant reduction in both groups. Most of the remaining clinical results showed stabilization or a small reduction (Table 8).

Secondary outcomes

Visual acuity BCVA

Accelerated corneal cross-linking

In all 4 trials^38,52,58,59 there was a significant increase in BCVA throughout the final follow-up (mean 14 months) with no intergroup difference. The majority of the other prospective studies showed a significant improvement (Tables 1–3). In pediatric cases^27,30,45 there was also a significant improvement. Studies that evaluated pulsed modified ACXL also found a significant improvement.

Transepithelial corneal cross-linking

All but one of the trials⁴⁶ showed a significant improvement in the BCVA. In the pediatric studies^63–65 half of them showed a statistical improvement and the other half, stabilization. Most of the clinical prospective studies showed a significant improvement (Tables 5–6).

Iontophoretic transepithelial corneal cross-linking

All the RCTs and most of the prospective studies showed a significant improvement of BCVA (Table 8).

Loss of vision (>1 line)

Accelerated corneal cross-linking

Most of the reports showed no vision loss; <10% of patients lost >1 line in a few studies^25,66,67 at 24 and 12 months of follow-up, respectively, except for Cınar et al.^29,57 who had 21% of patients that lost >1 line but the follow-up time was only 6 months.

Transepithelial corneal cross-linking

The majority of the clinical results showed no changes in BCVA, a few studies report some vision loss in <5% of patients^48,63 at 24 and 6 months of follow-up time, respectively.

Iontophoretic transepithelial corneal cross-linking

None of the reports showed vision loss during the follow-up.

Demarcation line depth

Not all the studies report this feature but it could be important because it works as an indirect marker of treatment efficacy. The demarcation line observed after the procedure is the theoretical limit between the anterior treated stroma and the posterior nontreated stroma. This is measured with the use of anterior segment tomography in the central cornea and in the periphery.

Accelerated corneal cross-linking

Studies have shown that the depth of the demarcation line in the central cornea is slightly >200 μm, whereas in the conventional protocol it is ∼300 μm.⁴⁴ Shetty et al.⁵⁹ was the only one using a 9 mW/cm² × 10 min procedure to achieve a depth (slightly <300 μm) similar to the conventional one. Protocols with shorter irradiation time seem to result in a shallower demarcation line than the conventional protocol.^{26,34,40,59,61} The 9 mW/cm² × 10 min⁶¹ and 18 mW/cm² × 5 min protocols⁵⁹ showed a demarcation line ∼200 μm. On the contrary, the 30 mW/cm² × 4 min protocols^35,59 showed the line <200 μm.

Transepithelial corneal cross-linking

Only a few studies report the depth of the demarcation line, which is slightly <200 μm or even not visible as described by Soeters et al.¹⁵

Iontophoretic transepithelial corneal cross-linking

Jouve et al.⁵⁴ found that the corneal demarcation line was visible by 1 month after surgery in 35% of cases with a mean depth of 216 μm, and Bikbova and Bikbov²⁰ showed a 170 μm depth at 2 weeks in half of the patients, which then disappeared.

Failure rate or progression

The definition of failure after CXL varies in the literature. A common variable in the definition is an increase of >0.50–1.0 diopters in 6–12 months of follow-up^15,43 or >1.0 diopter in maximum keratometry (Kmax) by some other studies.^{20,36,38,46,52,55,58,59}

Accelerated corneal cross-linking

These rates vary among the studies, from 3% found in an RCT by Shetty et al.⁵⁹ with 12 months of follow-up to 15.4% at 12 months in pediatric cases³⁰ and the majority are not reported at all.

Transepithelial corneal cross-linking

The failure rate was higher in this group. Two RCTs made by Soeters et al.¹⁵ and Al Fayez et al.⁴⁶ found 20% and 55% of progression at 12 and 36 months, respectively. Çerman et al.⁶⁸ and Caruso et al.³¹ reported ∼20% of failure in the TECXL group at 18 months of follow-up. In pediatric cases, Henriquez et al.⁶⁴ and Eraslan et al.⁶³ found ∼12% failure at 12 and 24 months, respectively.

Iontophoretic transepithelial corneal cross-linking

The failure rate found by Jouve et al.⁵⁴ was 20% and 18% by Lombardo et al.⁵⁵ at 24 and 12 months, respectively. Most of the studies did not report their failure rate.

Pulsed corneal cross-linking

We have to bear in mind that the biomechanical effect of CXL seems to be oxygen dependent. This dependency will be of particular importance in high fluence and TECXL and will most likely require major protocol modifications to maintain the efficiency of the method.³²

Mazzotta et al.³⁵ reported unchanged BCVA with stabilization of Kmax readings, and Moramarco et al.⁴⁰ retrospectively found a deeper demarcation line in the pulsed group; both authors used a 30 mW/cm² × 4 min protocol and 8 min for the pulsed procedure. Jiang et al.³⁴ compared it with the standard protocol achieving a significant improvement in both Kmax and BCVA; Artola et al.⁴² used a TECXL pulsed procedure in a PCS reporting stabilization in BCVA and Kmax, both authors followed them up for 12 months.

Conclusion

The new corneal collagen cross-linking protocols that reduce the exposure time and preserve the corneal epithelium without removal are safer approaches in terms of reducing complications and cause less visual loss compared with the standard Dresden approach. However, a lack of long-term studies and the different protocols used in accelerated and transepithelial methods and the failure rates reported by some studies make it difficult to provide definite conclusions regarding its real efficacy in halting the progression of corneal ectatic disorders.

The modeling effect on the cornea induced by these modified procedures varies considerably and the corneal flattening and the demarcation line are not as pronounced as with the standard Dresden corneal collagen cross-linking protocol. The shallower demarcation line could be the result of less effective accelerated or transepithelium methods. However, by treating the stroma farther from the endothelium, these modified procedures may be a safer option for treating thin corneas.

It seems that the future of corneal collagen cross-linking lies within the combination of accelerated and transepithelium protocols as published scientific studies and rationale have demonstrated its safety. Nevertheless, the technique should be standardized as currently there are many “recipes” with controversial results and further long-term studies including more patients are needed to demonstrate the efficacy of these protocols in halting the progression of corneal ectatic disorders.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

References

Georgiou

, Funnell

C.L.

, Cassels-Brown

, and O'Conor

Influence of ethnic origin on the incidence of keratoconus and associated atopic disease in Asians and white patients. Eye (Lond). 18:379–383, 2004.

Rabinowitz

Y.S.

Keratoconus. Surv. Ophthalmol. 42:297–319, 1998.

Gordon-Shaag

, Millodot

, Shneor

, and Liu

The genetic and environmental factors for keratoconus. Biomed Res. Int. 2015:795738, 2015.

Raiskup

, and Spoerl

Corneal crosslinking with riboflavin and ultraviolet A. I. Principles. Ocul. Surf., 11:65–74, 2013.

Wollensak

, Spoerl

, and Seiler

Riboflavin/ultraviolet-a-induced collagen crosslinking for the treatment of keratoconus. Am. J. Ophthalmol. 135:620–627, 2003.

Wittig-Silva

, Chan

, Islam

F.M.A.

, Wu

, Whiting

, and Snibson

G.R.

A randomized, controlled trial of corneal collagen cross-linking in progressive keratoconus. Ophthalmology, 121:812–821, 2014.

O'Brart

D.P.S.

, Chan

, Samaras

, Patel

, and Shah

S.P.

A randomised, prospective study to investigate the efficacy of riboflavin/ultraviolet A (370 nm) corneal collagen cross-linkage to halt the progression of keratoconus. Br. J. Ophthalmol. 95:1519–1524, 2011.

Abbouda

, Abicca

, and Alió

J.L.

Infectious keratitis following corneal crosslinking: a systematic review of reported cases: management, visual outcome, and treatment proposed. Semin. Ophthalmol. 31:485–491, 2016.

Abdelghaffar

, Hantera

, and Elsabagh

Corneal collagen cross-linking: promises and problems. Br. J. Ophthalmol. 94:1559–1560, 2010.

10.

Ashwin

P.T.

, and McDonnell

P.J.

Collagen cross-linkage: a comprehensive review and directions for future research. Br. J. Ophthalmol. 94:965–970, 2010.

11.

Dhawan

, Rao

, and Natrajan

Complications of corneal collagen cross-linking. J. Ophthalmol. 2011:869015, 2011.

12.

Torricelli

A.A.M.

, Ford

M.R.

, Singh

, Santhiago

M.R.

, Dupps

W.J.

, and Wilson

S.E.

BAC-EDTA transepithelial riboflavin-UVA crosslinking has greater biomechanical stiffening effect than standard epithelium-off in rabbit corneas. Exp. Eye Res. 125:114–117, 2014.

13.

Caporossi

, Mazzotta

, Paradiso

A.L.

, Baiocchi

, Marigliani

, and Caporossi

Transepithelial corneal collagen crosslinking for progressive keratoconus: 24-month clinical results. J. Cataract Refract. Surg. 39:1157–1163, 2013.

14.

Leccisotti

, and Islam

Transepithelial corneal collagen cross-linking in keratoconus. J. Refract. Surg. 26:942–948, 2010.

15.

Soeters

, Wisse

R.P.L.

, Godefrooij

D.A.

, Imhof

S.M.

, and Tahzib

N.G.

Transepithelial versus epithelium-off corneal cross-linking for the treatment of progressive keratoconus: a randomized controlled trial. Am. J. Ophthalmol. 159:821.e3–828.e3, 2015.

16.

Rechichi

, Daya

, Scorcia

, Meduri

, and Scorcia

Epithelial-disruption collagen crosslinking for keratoconus: one-year results. J. Cataract Refract. Surg. 39:1171–1178, 2013.

17.

Cassagne

, Laurent

, Rodrigues

, Galinier

, Spoerl

, Galiacy

S.D.

, et al. Iontophoresis transcorneal delivery technique for transepithelial corneal collagen crosslinking with riboflavin in a rabbit model. Invest. Ophthalmol. Vis. Sci. 57:594–603, 2016.

18.

Bikbova

, and Bikbov

Transepithelial corneal collagen cross-linking by iontophoresis of riboflavin. Acta Ophthalmol. 92:e30–e34, 2014.

19.

Vinciguerra

, Randleman

J.B.

, Romano

, Legrottaglie

E.F.

, Rosetta

, Camesasca

F.I.

, et al. Transepithelial iontophoresis corneal collagen cross-linking for progressive keratoconus: initial clinical outcomes. J. Refract. Surg. 30:746–753, 2014.

20.

Bikbova

, and Bikbov

Standard corneal collagen crosslinking versus transepithelial iontophoresis-assisted corneal crosslinking, 24 months follow-up: randomized control trial. Acta Ophthalmol. 94:e600–e606, 2016.

21.

Bunsen

R.W.

, and Roscoe

H.E.

Photochemical researches, part V: on the measurement of the chemical action of direct and diffuse sunlight. Proc. R. Soc. Lond. 12:306–312, 1862.

22.

Chan

T.C.Y.

, Chow

V.W.S.

, Jhanji

, and Wong

V.W.Y.

Different topographic response between mild to moderate and advanced keratoconus after accelerated collagen cross-linking. Cornea, 34:922–927, 2015.

23.

Ozgurhan

E.B.

, Akcay

B.I.S.

, Kurt

, Yildirim

, and Demirok

Accelerated corneal collagen cross-linking in thin keratoconic corneas. J. Refract. Surg. 31:386–390, 2015.

24.

Marino

G.K.

, Torricelli

A.A.M.

, Giacomin

, Santhiago

M.R.

, Espindola

, and Netto

M.V.

Accelerated corneal collagen cross-linking for postoperative LASIK ectasia: two-year outcomes. J. Refract. Surg. 31:380–384, 2015.

25.

Waszczykowska

, and Jurowski

Two-year accelerated corneal cross-linking outcome in patients with progressive keratoconus. Biomed Res. Int. 2015:325157, 2015.

26.

Ozgurhan

E.B.

, Kara

, Cankaya

K.I.

, Kurt

, and Demirok

Accelerated corneal cross-linking in pediatric patients with keratoconus: 24-month outcomes. J. Refract. Surg. 30:843–849, 2014.

27.

Shetty

, Nagaraja

, Jayadev

, Pahuja

N.K.

, Kurian Kummelil

, and Nuijts

R.M.M.A.

Accelerated corneal collagen cross-linking in pediatric patients: two-year follow-up results. Biomed Res. Int. 2014:894095, 2014.

28.

Elbaz

, Shen

, Lichtinger

, Zauberman

N.A.

, Goldich

, Chan

C.C.

, et al. Accelerated (9-mW/cm2) corneal collagen crosslinking for keratoconus—a 1-year follow-up. Cornea, 33:769–773, 2014.

29.

Cınar

, Cingü

A.K.

, Turkcu

F.M.

, Yüksel

, Sahin

, Yıldırım

, et al. Accelerated corneal collagen cross-linking for progressive keratoconus. Cutan. Ocul. Toxicol. 33:168–171, 2014.

30.

Baenninger

P.B.

, Bachmann

L.M.

, Wienecke

, Thiel

M.A.

, and Kaufmann

Pediatric corneal cross-linking: comparison of visual and topographic outcomes between conventional and accelerated treatment. Am. J. Ophthalmol. 183:11–16, 2017.

31.

Caruso

, Ostacolo

, Epstein

R.L.

, Barbaro

, Troisi

, and Capobianco

Transepithelial corneal cross-linking with vitamin E-enhanced riboflavin solution and abbreviated, low-dose UV-A: 24-month clinical outcomes. Cornea, 35:145–150, 2016.

32.

Richoz

, Hammer

, Tabibian

, Gatzioufas

, and Hafezi

The biomechanical effect of corneal collagen cross-linking (CXL) with riboflavin and UV-A is oxygen dependent. Transl. Vis. Sci. Technol. 2:6, 2013.

33.

Kling

, Richoz

, Hammer

, Tabibian

, Jacob

, Agarwal

, et al. Increased biomechanical efficacy of corneal cross-linking in thin corneas due to higher oxygen availability. J. Refract. Surg. 31:840–846, 2015.

34.

Jiang

L-Z

, Jiang

, and Qiu

S.-Y.

Conventional vs. pulsed-light accelerated corneal collagen cross-linking for the treatment of progressive keratoconus: 12-month results from a prospective study. Exp. Ther. Med., 14:4238–4244, 2017.

35.

Mazzotta

, Traversi

, Caragiuli

, and Rechichi

Pulsed vs continuous light accelerated corneal collagen crosslinking: in vivo qualitative investigation by confocal microscopy and corneal OCT. Eye, 28:1179–1183, 2014.

36.

Stojanovic

, Zhou

, and Utheim

T.P.

Corneal collagen cross-linking with and without epithelial removal: a contralateral study with 0.5% hypotonic riboflavin solution. Biomed Res. Int. 2014:619398, 2014.

37.

Hafezi

, Mrochen

, Iseli

H.P.

, and Seiler

Collagen crosslinking with ultraviolet-A and hypoosmolar riboflavin solution in thin corneas. J. Cataract Refract. Surg. 35:621–624, 2009.

38.

Hashemian

, Jabbarvand

, Khodaparast

, and Ameli

Evaluation of corneal changes after conventional versus accelerated corneal cross-linking: a randomized controlled trial. J. Refract. Surg. 30:837–842, 2014.

39.

Magli

, Chiariello Vecchio

, Carelli

, Piozzi

, Di Landro

, and Troisi

Pediatric keratoconus and iontophoretic corneal crosslinking: refractive and topographic evidence in patients underwent general and topical anesthesia, 18 months of follow-up. Int. Ophthalmol. 36:585–590, 2016.

40.

Moramarco

, Iovieno

, Sartori

, and Fontana

Corneal stromal demarcation line after accelerated crosslinking using continuous and pulsed light. J. Cataract Refract. Surg. 41:2546–2551, 2015.

41.

Zhang

, Sun

, Chen

, Li

, Tian

, and Zhou

One-year outcomes of pachymetry and epithelium thicknesses after accelerated (45 mW/cm(2)) transepithelial corneal collagen cross-linking for keratoconus patients. Sci. Rep. 6:32692, 2016.

42.

Artola

, Piñero

D.P.

, Ruiz-Fortes

, Soto-Negro

, and Pérez-Cambrodí

R.J.

Clinical outcomes at one year following keratoconus treatment with accelerated transepithelial cross-linking. Int. J. Ophthalmol. 10:652–655, 2017.

43.

Nawaz

, Gupta

, Gogia

, Sasikala

N.K.

, and Panda

Trans-epithelial versus conventional corneal collagen crosslinking: a randomized trial in keratoconus. Oman J. Ophthalmol. 8:9–13, 2015.

44.

Brittingham

, Tappeiner

, and Frueh

B.E.

Corneal cross-linking in keratoconus using the standard and rapid treatment protocol: differences in demarcation line and 12-month outcomes. Invest. Ophthalmol. Vis. Sci. 55:8371–8376, 2014.

45.

Badawi

A.E.

Accelerated corneal collagen cross-linking in pediatric keratoconus: one year study. Saudi J. Ophthalmol. 31:11–18, 2017.

46.

Al Fayez

M.F.

, Alfayez

, and Alfayez

Transepithelial versus epithelium-off corneal collagen cross-linking for progressive keratoconus: a prospective randomized controlled trial. Cornea, 34(Suppl 10):S53–S56, 2015.

47.

De Bernardo

, Capasso

, Tortori

, Lanza

, Caliendo

, and Rosa

Trans epithelial corneal collagen crosslinking for progressive keratoconus: 6 months follow up. Cont. Lens Anterior Eye, 37:438–441, 2014.

48.

Lesniak

S.P.

, and Hersh

P.S.

Transepithelial corneal collagen crosslinking for keratoconus: six-month results. J. Cataract Refract. Surg. 40:1971–1979, 2014.

49.

Koppen

, Wouters

, Mathysen

, Rozema

, and Tassignon

M.-J.

Refractive and topographic results of benzalkonium chloride-assisted transepithelial crosslinking. J. Cataract Refract. Surg. 38:1000–1005, 2012.

50.

Spadea

, and Mencucci

Transepithelial corneal collagen cross-linking in ultrathin keratoconic corneas. Clin. Ophthalmol. 6:1785–1792, 2012.

51.

Razmjoo

, Peyman

, Rahimi

, and Modrek

H.J.

Cornea collagen cross-linking for keratoconus: a comparison between accelerated and conventional methods. Adv. Biomed. Res. 6:10, 2017.

52.

Hashemi

, Miraftab

, Seyedian

M.A.

, Hafezi

, Bahrmandy

, Heidarian

, et al. Long-term results of an accelerated corneal cross-linking protocol (18 mW/cm²) for the treatment of progressive keratoconus. Am. J. Ophthalmol. 160:1164.e1–1170.e1, 2015.

53.

Rossi

, Orrico

, Santamaria

, Romano

, De Rosa

, Simonelli

, et al. Standard versus trans-epithelial collagen cross-linking in keratoconus patients suitable for standard collagen cross-linking. Clin. Ophthalmol. 9:503–509, 2015.

54.

Jouve

, Borderie

, Sandali

, Temstet

, Basli

, Laroche

, et al. Conventional and iontophoresis corneal cross-linking for keratoconus: efficacy and assessment by optical coherence tomography and confocal microscopy. Cornea, 36:153–162, 2017.

55.

Lombardo

, Giannini

, Lombardo

, and Serrao

Randomized controlled trial comparing transepithelial corneal cross-linking using iontophoresis with the Dresden protocol in progressive keratoconus. Ophthalmology, 124:804–812, 2017.

56.

Buzzonetti

, Petrocelli

, Valente

, Iarossi

, Ardia

, and Petroni

Iontophoretic transepithelial corneal cross-linking to halt keratoconus in pediatric cases: 15-month follow-up. Cornea, 34:512–515, 2015.

57.

Cınar

, Cingü

A.K.

, Türkcü

F.M.

, Çınar

, Yüksel

, Özkurt

Z.G.

, et al. Comparison of accelerated and conventional corneal collagen cross-linking for progressive keratoconus. Cutan. Ocul. Toxicol. 33:218–222, 2014.

58.

Sherif

A.M.

Accelerated versus conventional corneal collagen cross-linking in the treatment of mild keratoconus: a comparative study. Clin. Ophthalmol. 8:1435–1440, 2014.

59.

Shetty

, Pahuja

N.K.

, Nuijts

R.M.M

.A., Ajani

, Jayadev

, Sharma

, et al. Current protocols of corneal collagen cross-linking: visual, refractive, and tomographic outcomes. Am. J. Ophthalmol. 160:243–249, 2015.

60.

Sadoughi

M.M.

, Einollahi

, Baradaran-Rafii

, Roshandel

, Hasani

, and Nazeri

Accelerated versus conventional corneal collagen cross-linking in patients with keratoconus: an intrapatient comparative study. Int. Ophthalmol. 38:67–74, 2018.

61.

A.L.K.

, Chan

T.C.Y.

, and Cheng

A.C.K.

Conventional versus accelerated corneal collagen cross-linking in the treatment of keratoconus. Clin. Exp. Ophthalmol. 44:8–14, 2016.

62.

Kocak

, Aydin

, Kaya

, and Koc

Comparison of transepithelial corneal collagen crosslinking with epithelium-off crosslinking in progressive keratoconus. J. Fr. Ophtalmol. 37:371–376, 2014.

63.

Eraslan

, Toker

, Cerman

, and Ozarslan

Efficacy of epithelium-off and epithelium-on corneal collagen cross-linking in pediatric keratoconus. Eye Contact Lens, 43:155–161, 2017.

64.

Henriquez

M.A.

, Rodríguez

A.M.

, and Izquierdo

Accelerated epi-on versus standard epi-off corneal collagen cross-linking for progressive keratoconus in pediatric patients. Cornea, 36:1503–1508, 2017.

65.

Magli

, Forte

, Tortori

, Capasso

, Marsico

, and Piozzi

Epithelium-off corneal collagen cross-linking versus transepithelial cross-linking for pediatric keratoconus. Cornea, 32:597–601, 2013.

66.

Cummings

A.B.

, McQuaid

, Naughton

, Brennan

, and Mrochen

Optimizing corneal cross-linking in the treatment of keratoconus: a comparison of outcomes after standard- and high-intensity protocols. Cornea, 35:814–822, 2016.

67.

Yıldırım

, Olcucu

, Gunaydin

Z.K.

, Ağca

, Ozgurhan

E.B.

, Alagoz

, et al. Comparison of accelerated corneal collagen cross-linking types for treating keratoconus. Curr. Eye Res. 42:971–975, 2017.

68.

Çerman

, Toker

, and Ozarslan Ozcan

Transepithelial versus epithelium-off crosslinking in adults with progressive keratoconus. J. Cataract Refract. Surg. 41:1416–1425, 2015.

69.

Alnawaiseh

, Rosentreter

, Böhm

M.R.R.

, Eveslage

, Eter

, and Zumhagen

Accelerated (18 mW/cm²) corneal collagen cross-linking for progressive keratoconus. Cornea, 34:1427–1431, 2015.

70.

Kanellopoulos

A.J.

Long term results of a prospective randomized bilateral eye comparison trial of higher fluence, shorter duration ultraviolet A radiation, and riboflavin collagen cross linking for progressive keratoconus. Clin. Ophthalmol. 6:97–101, 2012.

71.

Tomita

, Mita

, and Huseynova

Accelerated versus conventional corneal collagen crosslinking. J. Cataract Refract. Surg. 40:1013–1020, 2014.

72.

Chow

V.W.S.

, Chan

T.C.Y.

, Yu

, Wong

V.W.Y.

, and Jhanji

One-year outcomes of conventional and accelerated collagen crosslinking in progressive keratoconus. Sci. Rep. 5:14425, 2015.

73.

Woo

J.H.

, Iyer

J.V.

, Lim

, Hla

M.H.

, Mehta

J.S.

, Chan

C.M.

, et al. Conventional versus accelerated collagen cross-linking for keratoconus: a comparison of visual, refractive, topographic and biomechanical outcomes. Open Ophthalmol. J. 11:262–272, 2017.

74.

Filippello

, Stagni

, and O'Brart

Transepithelial corneal collagen crosslinking: bilateral study. J. Cataract Refract. Surg. 38:283–291, 2012.

75.

Salman

A.G.

Transepithelial corneal collagen crosslinking for progressive keratoconus in a pediatric age group. J. Cataract Refract. Surg. 39:1164–1170, 2013.

76.

Khairy

H.A.

, Marey

H.M.

, and Ellakwa

A.F.

Epithelium-on corneal cross-linking treatment of progressive keratoconus: a prospective, consecutive study. Clin. Ophthalmol. 29;8:819–823, 2014.

77.

Akbar

, Intisar-Ul-Haq

, Ishaq

, Arzoo

, and Siddique

Transepithelial corneal crosslinking in treatment of progressive keratoconus: 12 months' clinical results. Pak. J. Med. Sci. 33:570–575, 2017.

78.

Aixinjueluo

, Usui

, Miyai

, Toyono

, Sakisaka

, and Yamagami

Accelerated transepithelial corneal cross-linking for progressive keratoconus: a prospective study of 12 months. Br. J. Ophthalmol. 101:1244–1249, 2017.

79.

Vinciguerra

, Romano

, Rosetta

, Legrottaglie

E.F.

, Piscopo

, Fabiani

, et al. Transepithelial iontophoresis versus standard corneal collagen cross-linking: 1-year results of a prospective clinical study. J. Refract. Surg. 32:672–678, 2016.

80.

Cantemir

, Alexa

A.I.

, Anton

, Ciuntu

R.E.

, Danielescu

, Chiselita

, et al. Evaluation of iontophoretic collagen cross-linking for early stage of progressive keratoconus compared to standard cross-linking: a non-inferiority study. Ophthalmol. Ther. 6:147–160, 2017.