Collagen Cross-Linking Using Riboflavin and Ultraviolet-A for Corneal Thinning Disorders: An Evidence-Based Analysis

G Pron; L Ieraci; K Kaulback; Medical Advisory Secretariat, Health Quality Ontario

. 2011 Nov 1;11(5):1–89.

Collagen Cross-Linking Using Riboflavin and Ultraviolet-A for Corneal Thinning Disorders

An Evidence-Based Analysis

G Pron, L Ieraci, K Kaulback; Medical Advisory Secretariat, Health Quality Ontario

PMCID: PMC3377552 PMID: 23074417

Executive Summary

Objective

The main objectives for this evidence-based analysis were to determine the safety and effectiveness of photochemical corneal collagen cross-linking with riboflavin (vitamin B₂) and ultraviolet-A radiation, referred to as CXL, for the management of corneal thinning disease conditions. The comparative safety and effectiveness of corneal cross-linking with other minimally invasive treatments such as intrastromal corneal rings was also reviewed. The Medical Advisory Secretariat (MAS) evidence-based analysis was performed to support public financing decisions.

Subject of the Evidence-Based Analysis

The primary treatment objective for corneal cross-linking is to increase the strength of the corneal stroma, thereby stabilizing the underlying disease process. At the present time, it is the only procedure that treats the underlying disease condition. The proposed advantages for corneal cross-linking are that the procedure is minimally invasive, safe and effective, and it can potentially delay or defer the need for a corneal transplant. In addition, corneal cross-linking does not adversely affect subsequent surgical approaches, if they are necessary, or interfere with corneal transplants. The evidence for these claims for corneal cross-linking in the management of corneal thinning disorders such as keratoconus will be the focus of this review.

The specific research questions for the evidence review were as follows:

Technical: How technically demanding is corneal cross-linking and what are the operative risks?
Safety: What is known about the broader safety profile of corneal cross-linking?
Effectiveness - Corneal Surface Topographic Affects:
1. What are the corneal surface remodeling effects of corneal cross-linking?
2. Do these changes interfere with subsequent interventions, particularly corneal transplant known as penetrating keratoplasty (PKP)?
Effectiveness -Visual Acuity:
1. What impacts does the remodeling have on visual acuity?
2. Are these impacts predictable, stable, adjustable and durable?
Effectiveness - Refractive Outcomes: What impact does remodeling have on refractive outcomes?
Effectiveness - Visual Quality (Symptoms): What impact does corneal cross-linking have on vision quality such as contrast vision, and decreased visual symptoms (halos, fluctuating vision)?
Effectiveness - Contact lens tolerance: To what extent does contact lens intolerance improve after corneal cross-linking?
Vision-Related QOL: What is the impact of corneal cross-linking on functional visual rehabilitation and quality of life?
Patient satisfaction: Are patients satisfied with their vision following the procedure?
Disease Process:
1. What impact does corneal cross-linking have on the underling corneal thinning disease process?
Does corneal cross-linking delay or defer the need for a corneal transplant?
What is the comparative safety and effectiveness of corneal cross-linking compared with other minimally invasive treatments for corneal ectasia such as intrastromal corneal rings?

Clinical Need: Target Population and Condition

Corneal ectasia (thinning) disorders represent a range of disorders involving either primary disease conditions, such as keratoconus (KC) and pellucid marginal corneal degeneration, or secondary iatrogenic conditions, such as corneal thinning occurring after laser in situ keratomileusis (LASIK) refractive surgery.

Corneal thinning is a disease that occurs when the normally round dome-shaped cornea progressively thins causing a cone-like bulge or forward protrusion in response to the normal pressure of the eye. The thinning occurs primarily in the stroma layers and is believed to be a breakdown in the collagen process. This bulging can lead to irregular astigmatism or shape of the cornea. Because the anterior part of the cornea is responsible for most of the focusing of the light on the retina, this can then result in loss of visual acuity. The reduced visual acuity can make even simple daily tasks, such as driving, watching television or reading, difficult to perform.

Keratoconus is the most common form of corneal thinning disorder and involves a noninflammatory chronic disease process of progressive corneal thinning. Although the specific cause for the biomechanical alterations in the corneal stroma is unknown, there is a growing body of evidence suggesting that genetic factors may play an important role. Keratoconus is a rare disease (< 0.05% of the population) and is unique among chronic eye diseases because it has an early onset, with a median age of 25 years. Disease management for this condition follows a step-wise approach depending on disease severity. Contact lenses are the primary treatment of choice when there is irregular astigmatism associated with the disease. Patients are referred for corneal transplants as a last option when they can no longer tolerate contact lenses or when lenses no longer provide adequate vision.

Keratoconus is one of the leading indications for corneal transplants and has been so for the last 3 decades. Despite the high success rate of corneal transplants (up to 20 years) there are reasons to defer it as long as possible. Patients with keratoconus are generally young and a longer-term graft survival of at least 30 or 40 years may be necessary. The surgery itself involves lengthy time off work and postsurgery, while potential complications include long-term steroid use, secondary cataracts, and glaucoma. After a corneal transplant, keratoconus may recur resulting in a need for subsequent interventions. Residual refractive errors and astigmatism can remain challenges after transplantation, and high refractive surgery and regraft rates in KC patients have been reported. Visual rehabilitation or recovery of visual acuity after transplant may be slow and/or unsatisfactory to patients.

Description of Technology/Therapy

Corneal cross-linking involves the use of riboflavin (vitamin B₂) and ultraviolet-A (UVA) radiation. A UVA irradiation device known as the CXL® device (license number 77989) by ACCUTECH Medical Technologies Inc. has been licensed by Health Canada as a Class II device since September 19, 2008. An illumination device that emits homogeneous UVA, in combination with any generic form of riboflavin, is licensed by Health Canada for the indication to slow or stop the progression of corneal thinning caused by progressive keratectasia, iatrogenic keratectasia after laser-assisted in situ keratomileusis (LASIK) and pellucid marginal degeneration. The same device is named the UV-X® device by IROCMedical, with approvals in Argentina, the European Union and Australia.

UVA devices all use light emitting diodes to generate UVA at a wavelength of 360-380 microns but vary in the number of diodes (5 to 25), focusing systems, working distance, beam diameter, beam uniformity and extent to which the operator can vary the parameters. In Ontario, CXL is currently offered at over 15 private eye clinics by refractive surgeons and ophthalmologists.

The treatment is an outpatient procedure generally performed with topical anesthesia. The treatment consists of several well defined procedures. The epithelial cell layer is first removed, often using a blunt spatula in a 9.0 mm diameter under sterile conditions. This step is followed by the application of topical 0.1% riboflavin (vitamin B₂) solution every 3 to 5 minutes for 25 minutes to ensure that the corneal stroma is fully penetrated. A solid-state UVA light source with a wavelength of 370 nm (maximum absorption of riboflavin) and an irradiance of 3 mW/cm² is used to irradiate the central cornea. Following treatment, a soft bandage lens is applied and prescriptions are given for oral pain medications, preservative-free tears, anti-inflammatory drops (preferably not nonsteroidal anti-inflammatory drugs, or NSAIDs) and antibiotic eye drops. Patients are recalled 1 week following the procedure to evaluate re-epithelialization and they are followed-up subsequently.

Evidence-Based Analysis Methods

A literature search was conducted on photochemical corneal collagen cross-linking with riboflavin (vitamin B₂) and ultraviolet-A for the management of corneal thinning disorders using a search strategy with appropriate keywords and subject headings for CXL for literature published up until April 17, 2011. The literature search for this Health Technology Assessment (HTA) review was performed using the Cochrane Library, the Emergency Care Research Institute (ECRI) and the Centre for Reviews and Dissemination. The websites of several other health technology agencies were also reviewed, including the Canadian Agency for Drugs and Technologies in Health (CADTH) and the United Kingdom’s National Institute for Clinical Excellence (NICE). The databases searched included OVID MEDLINE, MEDLINE IN-Process and other Non-Indexed Citations such as EMBASE.

As the evidence review included an intervention for a rare condition, case series and case reports, particularly for complications and adverse events, were reviewed. A total of 316 citations were identified and all abstracts were reviewed by a single reviewer for eligibility. For those studies meeting the eligibility criteria, full-text articles were obtained. Reference lists were also examined for any additional relevant studies not identified through the search.

Inclusion Criteria

English-language reports and human studies
patients with any corneal thinning disorder
reports with CXL procedures used alone or in conjunction with other interventions
original reports with defined study methodology
reports including standardized measurements on outcome events such as technical success, safety effectiveness, durability, vision quality of life or patient satisfaction
systematic reviews, meta-analyses, randomized controlled trials, observational studies, retrospective analyses, case series, or case reports for complications and adverse events

Exclusion Criteria

nonsystematic reviews, letters, comments and editorials
reports not involving outcome events such as safety, effectiveness, durability, vision quality or patient satisfaction following an intervention with corneal implants
reports not involving corneal thinning disorders and an intervention involving CXL

Summary of Evidence Findings

In the Medical Advisory Secretariat evidence review on corneal cross-linking, 65 reports (16 case reports) involving 1403 patients were identified on the use of CXL for managing corneal thinning disorders. The reports were summarized according to their primary clinical indication, whether or not secondary interventions were used in conjunction with CXL (referred to as CXL-Plus) and whether or not it was a safety-related report.

The safety review was based on information from the cohort studies evaluating effectiveness, clinical studies evaluating safety, treatment response or recovery, and published case reports of complications. Complications, such as infection and noninfectious keratitis (inflammatory response), reported in case reports, generally occurred in the first week and were successfully treated with topical antibiotics and steroids. Other complications, such as the cytotoxic effects on the targeted corneal stroma, occurred as side effects of the photo-oxidative process generated by riboflavin and ultraviolet-A and were usually reversible.

The reports on treatment effectiveness involved 15 pre-post longitudinal cohort follow-up studies ranging from follow-up of patients’ treated eye only, follow-up in both the treated and untreated fellow-eye; and follow-up in the treated eye only and a control group not receiving treatment. One study was a 3-arm randomized control study (RCT) involving 2 comparators: one comparator was a sham treatment in which one eye was treated with riboflavin only; and the other comparator was the untreated fellow-eye. The outcomes reported across the studies involved statistically significant and clinically relevant improvements in corneal topography and refraction after CXL. In addition, improvements in treated eyes were accompanied by worsening outcomes in the untreated fellow-eyes. Improvements in corneal topography reported at 6 months were maintained at 1- and 2-year follow-up. Visual acuity, although not always improved, was infrequently reported as vision loss. Additional procedures such as the use of intrastromal corneal ring segments, intraocular lenses and refractive surgical practices were reported to result in additional improvements in topography and visual acuity after CXL.

Considerations for Ontario Health System

The total costs of providing CXL therapy to keratoconus patients in Ontario was calculated based on estimated physician, clinic, and medication costs. The total cost per patient was approximately $1,036 for the treatment of one eye, and $1,751 for the treatment of both eyes. The prevalence of keratoconus was estimated at 4,047 patients in FY2011, with an anticipated annual incidence (new cases) of about 148 cases. After distributing the costs of CXL therapy for the FY2011 prevalent keratoconus population over the next 3 years, the estimated average annual cost was approximately $2.1 million, of which about $1.3 million would be physician costs specifically.

Conclusion

Corneal cross-linking effectively stabilizes the underlying disease, and in some cases reverses disease progression as measured by key corneal topographic measures. The affects of CXL on visual acuity are less predictable and the use of adjunct interventions with CXL, such as intrastromal corneal ring segments, refractive surgery, and intraocular lens implants are increasingly employed to both stabilize disease and restore visual acuity. Although the use of adjunct interventions have been shown to result in additional clinical benefit, the order, timing, and risks of performing adjunctive interventions have not been well established.

Although there is potential for serious adverse events with corneal UVA irradiation and photochemical reactions, there have been few reported complications. Those that have occurred tended to be related to side effects of the induced photochemical reactions and were generally reversible. However, to ensure that there are minimal complications with the use of CXL and irradiation, strict adherence to defined CXL procedural protocols is essential.

Keywords

Keratoconus, corneal cross-linking, corneal topography, corneal transplant, visual acuity, refractive error.

Background

Objective of Analysis

The overall scope of the project was to determine the role of photochemical corneal collagen cross-linking with riboflavin (vitamin B₂) and ultraviolet-A radiation, referred to as CXL, for the management of corneal thinning disease conditions. The main objectives for the evidence review were to determine the safety and effectiveness of CXL for the management of corneal thinning disorders. The comparative safety and effectiveness of CXL with other minimally invasive treatments such as intrastromal corneal rings was also reviewed.

Clinical Need

Corneal ectasia (thinning) disorders represent a range of disorders and can involve either primary disease conditions such as keratoconus (KC) and pellucid marginal corneal degeneration (PMCD) or secondary iatrogenic conditions such as corneal ectasia occurring after laser in-situ keratomileusis (LASIK) refractive surgery. Corneal thinning is a disease that occurs when the normally round dome-shaped cornea (the clear outer area of the eye) progressively thins, causing a cone-like bulge or forward protrusion in response to the normal pressure of the eye pushing out on the thinned areas of the cornea. (1) The thinning occurs primarily in the stroma layers and is believed to be a breakdown in the collagen process.

This bulging can lead to an irregularly shaped cornea and because the anterior part of the cornea is responsible for most of the focusing of the light on the retina, can results in loss of visual acuity, both uncorrected visual acuity (UCVA) and best-spectacle corrected visual acuity (BSCVA). The visual acuity loss is secondary to high irregular astigmatism that can occur with and without myopia. (2) The reduced visual acuity can make even simple daily tasks, such as driving, watching television or reading, difficult to perform. The subsequent corneal protrusion or distortions can also result in corneal scarring and treatment-related sequelae such as abrasions from contact lenses.

There are a variety of corneal thinning disorders but it is unknown if these represent distinct forms of the disease or variants of the same disease process. Keratoconus is the most common form of thinning disorders and involves a noninflammatory chronic disease process of progressive corneal thinning. (3) Although the condition may initially present in one eye, it is a progressive disorder and eventually affects both eyes. (4) In keratoconus, localized thinning can occur in a variety of patterns, but when it occurs in the inferior cornea in a crescent-type pattern it is referred to as pellucid marginal corneal degeneration (PMCD). (5)

Etiology

Keratoconus leads to biomechanical alterations of the cornea involving the collagen scaffold and collagen compound, and to their bonding with the collagen fibrils. (6;7) The biochemical resistance of the cornea of KC patients is half that of the normal value. (8) The specific cause for these biochemical alterations, however, is unknown. There is a growing body of evidence suggesting that genetic factors may play an important role. (4) The proportion of persons with KC reported to have a positive family history of the disease ranged from 6% to 15%. The National Eye Institute (NEI) in the United States sponsored a multicenter longitudinal follow-up cohort study, the Collaborative Longitudinal Evaluation of Keratoconus (CLEK), which found that out of 1,209 keratoconus patients recruited over a 1-year period (May 1995 to June 1996), 13.5% reported a family history of KC. (2) Also reported has been an occurrence of the disease in second- and third-generation studies and a high concordance in monozygotic twins. (9;10) A study evaluating corneal topography in first-degree relatives of patients with KC found an 11% (8/72) incidence of the disease, compared to 0.05% in the general population. (11)

Additional information about KC patients is available from the CLEK study, the largest cohort study of keratoconus patients to date. (2) A reported high percentage of atopia (53%) has unknown clinical significance for these patients. Unlike findings in smaller clinical series, no patients reported systemic diseases such as Down’s syndrome, Marfan syndrome, focal dermal hypoplasia, Ehlers-Danlos syndrome, oculodentodigital syndrome, osteogenesis imperfecta, or Reigers anomaly.

Diagnosis

Diagnosis of KC depends on the methods used and can be difficult for several reasons. The onset of the disease itself is gradual and in some patients may never progress beyond subtle irregular astigmatism for which patients may not seek medical or optometric care. Subclinical forms (forme fruste) of keratoconus are particularly difficult to diagnose. Representative diagnostic patterns for subclinical forms have been described by some as having a corneal topographic pattern of at least one of the following: Inferior-Superior (I-S) asymmetry index> 1.4 diopters (D), central corneal power > 47.2 D, or a fellow-eye diagnosed with keratoconus. (12)

Computer-assisted videophotokeratoscopes provides a means to detect subtle changes and quantitative measures of corneal surface topographic changes. (13) The most commonly employed grading or classification system for KC is Amsler-Krumeich scale. (3;14) This classification system comprises 4 stages of disease based on the degree of corneal topography, myopia or induced astigmatism, clinical signs (Vogl’s striae, etc), central corneal scarring, and corneal thickness. The disease stages are as follows:

Stage 1 - eccentric corneal steepening, induced myopia and/or astigmatism <5 D, corneal radii ≤ 48 D, slit lamp findings (Vogl’s striae), no scars
Stage 2 - induced myopia and/or astigmatism >5 D <8 D, corneal radii ≤ 53 D, no central scars, corneal thickness ~400 µm
Stage 3 - induced myopia and/or astigmatism >8 D <10 D, corneal radii >53 D, no central scars, corneal thickness 200 to 400 µm
Stage 4 - refraction not measurable, corneal radii >55D, central scars, perforation, corneal thickness 200 µm.

Disease prevalence and natural history

Keratoconus is a rare disorder with estimates of prevalence ranging from a rate of approximately 50 to 230 per 100,000 population. (15) An American population-based 48-year survey estimated an overall prevalence of 54.5 per 100,000, with an overall annual incidence rate of 2.0 per 100,000. (16) The age-adjusted prevalence rate was significantly (P <.05) higher in males than in females (69.5 vs. 39.2 per 100,000).

Keratoconus is unique in chronic eye diseases because it has an early onset. (17) In the Kennedy et al. study of 64 cases, the median age at disease onset was reported to be 25 years (ranging from 12 to 77 years) and the diagnosis was unilateral in 41% at diagnosis. In the CLEK study, the median age at study entry was 39.3 years and the impact of KC on vision was already detectable. (2) The variable visual acuity of KC patients in this study (Table 1) shows that 22% already had fair or worse (≥ 20/40) best-corrected visual acuity (BCVA) in their worst eye. A Snellen visual acuity range of up to 20/40 is interpreted as a range in which many individuals can function without optical correction. (18) A visual acuity above 20/40 is the most common cut-off level for unrestricted drivers’ licences, and 20/200 or worse is part of the legal definition of blindness.

Table 1: Baseline Visual Acuity of Keratoconus Patients in the CLEK Cohort Study^*.

Vision Quality	Best Corrected Visual Acuity Range	Snellen Visual Acuity Better Eye, No. (%)	Snellen Visual Acuity Worse Eye, No. (%)
Normal range	20/20 or better	538 (44.7)	169 (14.0)
Normal (without optical correction)	20/21 to 20/40	612 (50.8)	769 (63.9)
Fair	20/40 to 20/69	43 (3.6)	183 (15.2)
Poor	20/70 to 20/199	10 (0.8)	71 (5.9)
Poor (legal definition blindness)	20/200 or worse	1 (0.08)	12 (0.9)

Type of Correction	Number of Patients, (%)
Same in Both Eyes
Unaided	43 (3.6)
Glasses	194 (16.1)
Contact lenses	321 (26.6)
Glasses and contacts	571 (47.2)
Different in Each Eye
One eye unaided and fellow-eye contact lenses	37 (3.1)
One eye glasses and contacts with fellow-eye unaided	2 (0.2)
One eye glasses and contacts with fellow-eye contact lenses	1 (0.1)
One eye glasses and contacts with fellow-eye glasses	40 (3.6)

Patients With Ocular Disorders, N	Vision Range In Better Seeing Eye	Mean Utility Value	Standard Deviation	95% Confidence Interval
127	20/20 to 20/25	0.88	0.15	0.85−0.91
218	20/30 to 20/50	0.81	0.21	0.78−0.84
83	20/60 to 20/100	0.72	0.21	0.67−0.77
72	20/200 to no light perception	0.61	0.19	0.57−0.65

Disease State/Event	Time Tradeoff Utility Value
Diabetes	0.88
Visual acuity 20/40 (most common cut off for driver’s licence)	0.80
Myocardial infarction, moderate	0.80
Stroke, moderate (requiring some help but able to walk)	0.69
Visual acuity 20/200 (definition of legal blindness)	0.66
Osteoarthritis hip, mild	0.69
Ulcerative colitis, Preoperative	0.58
Renal disease, end-stage, home dialysis	0.49
Total blindness (no light perception)	0.26
Stroke, severe (total paralysis)	0.30

Author, Year	Site Country	Study Design	Study Population	Study Follow-up
Ghanem R, 2010 (32)	Department of Ophthalmology University of San Paulo and Sadalla Amin Ghanem Eye Hospital, Santa Catarina, Brazil	Case series	14 p (6 M, 8F) − 14 e Mean age 71.14 yrs± 11.7 (range 53 to 89 yrs) Bullous keratopathy	6 months
Iseli H, 2008 (29)	Institute for Refractive and Ophthalmic Surgery, Zurich, Switzerland	Case series	5 p (2 M, 3 F) Age range 27−66 years Therapy resistant infectious keratitis associated corneal melting, 4 with prior LASIK	Range 1−9 months
Makdoumi K, 2010 (30)	Department Ophthalmology, Orebro University Hospital, Orebro, Sweden	Case series	6 p − 7 e Severe infectious keratitis and corneal melting in all cases	Range 1−6 months
Mazzotta C, 2011 (34)	Department Ophthalmology, Siena University, Italy	Case report	38-year-old male Visual fluctuations and declining visual acuity, after radical keratotomy for KC-related myopia	12−months
Moren H, 2010 (31)	Department Ophthalmology, Regional Hospital Vasteras, Sweden	Case report	25-year old female Severe keratitis	9 months
Wollensak G, 2009 (33)	Eye Laser Institute, Department of Ophthalmology, Martin Luther University, Germany	Case reports	3 cases (3 e) Bullous keratopathy due to 3 different conditions-pseudophakia, corneal transplant rejection and Fuch’s endothelial dystrophy	8 months

High	Further research is very unlikely to change our confidence in the estimate of effect.
Moderate	Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low	Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low	Any estimate of effect is very uncertain.

Effectiveness-Keratoconus	Longitudinal Pre-Post Cohorts	6 reports	264 patients
	Comparative Pre-Post Cohorts – Untreated	5 reports	145 patients
	Comparative Pre-Post Cohort – Control Group	1 report	10 patients
	RCT (KC and Post-Lasik ectasia)	1 report	58 patients
Effectiveness-Post-Lasik-Ectasia	Longitudinal Pre-Post Cohorts	4 reports	66 patients
CXL-Plus: (CXL +ICRS)	Longitudinal Pre-Post Cohorts	6 reports	91 patients
CXL-Plus: (CXL + PRK)	Longitudinal Pre-Post Cohorts	8 reports	74 patients
CXL-Plus: (CXL +IOL)	Case Reports	2 reports	12 patients
Safety	Safety and Recovery Series	18 reports	668 patients
Safety	Complication Case reports	14 reports	15 patients

Study Design	Number of Eligible Studies
RCT Studies
Systematic review of RCTs
Large RCT
Small RCT	4
Observational Studies
Systematic review of non-RCTs with contemporaneous controls	2
Non-RCT with contemporaneous controls	6
Systematic review of non-RCTs with historical controls
Non-RCT with historical controls	37
Database, registry, or cross-sectional study
Case reports	16
Retrospective review, modeling
Studies presented at an international conference or other sources of grey literature
Expert opinion

Author, Year	Patients (eyes) Mean Age ± SD	Keratometry (K)	Pre-post 12 mo Change Mean Diopter ± SD	Pre-post P -Value
Agrawal V, 2009 (49)	25 P (37 e) 16.9 ± 3.5 yrs	K-Max K-Apex	−2.47 ± 3.89 D −2.73 ± 7.95 D	0.004 0.004
Arbelaez M, 2009 (50)	19 P (20 e) 24.4 yrs (R 18 to 44)	K- Average K- Apex	−1.36 D −1.40 D	0.004 0.01
Doors M, 2009 (52)	29 P (29 e) 35.1±11.7 yrs	Central K K-Max	0.64 ± 1.73 D − 0.29 ± 2.05 D	> 0 .05 > 0.05
El-Raggal T, 2009 (53)	9 P (15 e) 26.4 yrs (R 21 to 31)	K-Average K-Max	−1.50 D −1.75 D	< 0.001 < 0.001
Raiskup-Wolf F, 2008 (6)	130 P (142 e) 30 ± 10.5 yrs	K-Apex K-Max	−2.68± 7.61 D −1.46 ± 3.76 D	< 0. 01 < 0.01
Saffarian L, 2010 (54)	53 P (92 e) 21.5 ±3.4 yrs	Sim-K	−0.94 ± 0.71 D	< 0 .001

Author, Year	Patients (eyes) Mean Age	Keratometry (K)	Treated Eye Pre-post 12 Mo Change Mean Diopter ± SD P -Value	Untreated Eye Pre-post 12 Mo Change Mean Diopter ± SD
Caporossi A, 2010 (41)	44 Patients (44 e) (R, 10 to 40 yrs)	K-Average	−1.96 ± 0.63 D P< 0. 05	+1.2 ± 0.96 D
Henriquez M, 2011 (44)	10 Patients (10 e), 10 control subjects 29.7 (R,15 to 43) yrs	K-Max K- Min	−2.66 D P = 0.04 −1.61 D P= 0.03	90% (9/10) increased K-Max
Koller T, 2009 (45)	21 Patients (21 e) Age NR	Minimum corneal curvature radius	55.0 to 54.3D P= 0 .01	48.6 to 49.2 D P = 0.002
Leccisotti A, 2010 (46)	51 Patients (51 e) 26.9 ± 6.3 yrs	K-Apex Sim-K	− 0.51 ± 7.79 P > 0.05 −0.10 ± 1.44 P> 0.05	1.61 ± 6.28 D P > 0.05 0.88 ± 2.35 D P > 0.05
Vinciguerra P, 2009 (42)	28 Patients (28 e) (R, 24 to 52 yrs)	K-Min K-Max	46.10 to 40.22 D P= 0 .0003 50.37 to 44.21 D P= 0.0011	NR
Wollensak G, 2003 (7)	22 Patients (23 e) 31.7 ±11.9	K-Max	2.01 ± 1.74 D P= 0.03	NR

Author Year	Patients (P) Mean Age± SD	Topographic Measure	Pre-Post Change	P -Value
Agrawal V, 2009 (49)	25 P (37 e) 16.9 ± 3.5 years	Corneal wavefront surface aberrometry	Spherical and higher order corneal aberrations did not show significant changes	> 0 .05
Agrawal V, 2009 (49)	25 P (37 e) 16.9 ± 3.5 years	Corneal wavefront surface aberrometry	Coma component (lower order aberrations) showed significant reduction	0.003
Arbelaez M, 2009 (50)	19 P (20 e) 24.4 years	Corneal wavefront surface aberrometry	Spherical and higher order corneal aberrations did not show significant changes	0.041
Arbelaez M, 2009 (50)	(R 18 to 44 yrs)	Corneal wavefront surface aberrometry	Absolute RMS and absolute coma were significantly reduced	0.026
Caporossi A, 2010 (41)	44 P (44 e) (R 10 to 40 yrs)	Corneal wavefront surface aberrometry Surface aberrometry (CSO Eye Top)	Total wavefront higher order aberrations were significantly reduced	< 0 .00001
		Corneal symmetry-Inferior-Superior-inferior index (SI)	Pre-post SI asymmetry index significantly improved	< 0.0001
		Corneal symmetry-Inferior-Superior-inferior index (SI)	Spherical aberration remained unchanged	> 0 .05
Doors M, 2009 (52)	29 P (29 e) 35.1 ±11.7 years (R 19 to 76 yrs)	Corneal aberrometry (IRX-3 Wavefront Aberrometer)	Higher-order aberration values coma-x, coma-y and spherical aberration were not significantly changed	> 0.05

Author, Year	Eyes	Mean Refractive Sphere (Mean Diopters ± SD)
Author, Year	Eyes	Preop	Postop	Pre-Post Change1.20 D in 47%	P - Value
Agrawal V, 2009 (49)	37	−7.24 ± 4.67 D	NR	1.20 D in 47%	0 .005
Arbelaez M, 2009 (50)	20	−3.84 ± 5.10 D	−2.58 ± 3.22 D	NR	0 .033
Caporossi A, 2006 (41)	44	NR	NR	1.62 ± 1.03 D (R 0 to 3.75 D)	< 0.00001
El-Raggal T, 2009 (53)	15	−3.20 ± 1.46 D	−2.73 ± 1.56 D	NR	< 0 .001
Saffarian L, 2010 (54)	92	−1.06 ± 1.92 D	−0.87 ± 1.60 D	−0.18 ± 0.790 D	> 0 .05
Vinciguerra P, 2009 (42)	28	−1.86 ± 2.58 D	−1.58 ± 2.64 D	NR	> 0 .05

Author, Year	Eyes	Mean Refractive Cylinder (Mean Diopters ± SD)
Author, Year	Eyes	Preop	Postop	Pre-Post Change	P - Value
Arbelaez M, 2009 (50)	20	−4.04 ± 1.52 D	−2.79 ± 1.13 D	NR	0.0003
Caporossi A, 2006 (41)	44	NR	NR	−0.52 ± 0.38 D R 0.75 to −2.0 D	> 0.05
El-Raggal T, 2009 (53)	15	4.90 ± 0.74 D R 3.75 to 6.0 D	4.95 ± 0.76 D R 4.0 to 6.0 D	NR	0.384
Saffarian L, 2010 (54)	92	−3.93 ± 1.67 D	−3.14 ± 1.50 D	0.78 ± 1.49 D	< 0.001
Vinciguerra P, 2009 (42)	28	−302 ± 174 D	−2.76 ± 1.11 D		< 0.05

Author, Year	Patients	Improvement BCVA ≥ 1Snellen line	No Change BCVA	Loss BCVA
Koller T, 2009 (55)	105	NR	NR	2.9% (95% CI 0.6% - 8.5%) ≥ 2 Snellen line
Agrawal V, 2009 (49)	37	20 (54%)	10 (28%) Lines Not defined	NR
Arbelaez M, 2009 (50)	20	12 (60%)	8 (40%) Lines Not defined	−
Doors M, 2009 (52)	20	10 (50%)	5 (25%) Lines Not defined	NR
El-Raggal T, 2009 (53)	15	4 (27%)	NR	1 (7%) 1 line
Raiskup-Wolf F, 2008 (6)	142	75 (53%)	29 (20%) Lines Not defined	NR
Hersch P, 2011 (47)	49	22 (45%)	23 (46%) Within 1 Snellen line	4 (10%) 1 Snellen line
Wollensak G, 2003 (7)	22	11 (50%)	9 (41%) Within 1 Snellen line	2 (9%) ≥ 1 Snellen line
Range 27%		27% − 60%	20% − 46%	2.9% − 10%

Author, Year	Patients	Regression Decrease K-Max	Stable K-Max	Progression Increase K-Max
Koller T, 2009 (55)	105	39 (37%) > 1 D	58 (55%) ± 1 D	8 (8%) > 1 D
Agrawal V, 2009 (49)	37	20 (54%) > 1 D	14 (38%) ± 0.5 D	3 (8%) > 1 D
Raiskup-Wolf F, 2008 (6)	142	80 (56%) > 1 D	43 (30%) ± 0.5 D	NR
Hersch P, 2011 (47)	49	17 (35%)> 2 D	NR	5 (10%) > 1 D
Wollensak G, 2003 (7)	22	10 (45%) > 2 D 12 (55%) >1 D	8 (36%) ±1.0 D	0
Range		35% − 56%	30% − 55%	8% − 10%

Author Year	Patients (P) Mean Age ± SD	Topographic measure	Treated Eye Pre-Post Change	Untreated Fellow-Eye Pre-Post Change	Treated versus Untreated Pre-Post Change	P-Value
Koller T, 2009 (45)	21 P (21 e) Age NR	Corneal topography Pentacam system with Scheimpflug camera – Minimum corneal curvature radius (Rmin) At 12-months	Rmim/mm 6.14 to 6.21 P= 0.01 Rmin/D 55.0 D to 54.3D P= 0.01 4 / 7 Pentacam KC indices showed significant improvements towards a more normalized corneal anterior surface Based on Rmin Change 0.12 mm ≃ 1 D Progressed: N = 0 Unchanged: N = 13 Regressed: N = 8	Rmin/mm 6.94 to 6.86 P= 0.002 Rmin/D48.6 D to 49.2 D P= 0.002 Based on Rmin Change 0.12 mm ≃ 1 D Progressed: N = 7 Unchanged: N = 14 Regressed: N = 0	Rmin/mm 0.066 ± 0.10 vs -0.08 ± 0.10 Rmin/D-0.62 ± 0.9 D vs 0.57 ± 0.8 D	0.0009 0.0009
Leccisotti A, 2010 (46)	51 P (51 e) 26.9 ± 6.3 (R 26.9 to 39) years	Corneal surface regularity index (ISV abnormal >37) At 12-months Tangential video keratography (Keratograph)	ISV 76.4 ± 28.2 to 77.3 ± 28.8 Mean change 0.9 ± 4.69 (p>.05)	ISV 48.2 ± 14.3 to 53.5 ± 19.5 Mean change 5.3 ± 7.30 (p>.05)	ISV (95% CI) 1.99 to 6.81
Vinciguerra P, 2009 (42)	28 P (28 e) R 24 to 52 years	Total (corneal and internal) wavefront analysis performed with Nidek OPD-Scan – 21-Klyce Indices	At 12-months, the Klyce indices had significantly improved (p<.05), 19 of the 21 indices improved, 1 remained the same (IAI),	At 12-months, the Klyce indices were significantly worse, none of the indices improved, 17 worsened, and 4 remained the same
			In wavefront analysis, total (corneal and internal) aberrations, total higher order aberrations, total astigmatism, total coma, and total spherical aberrations were significantly decreased i.e., improved	In wavefront analysis, total astigmatism and total coma significantly increased, i.e., worsened. No significant change was noted with total, higher order or spherical aberrations

Author, Year Patient (eyes) Mean Age	Observation Point	Topography K-Max (Mean Diopter ± SD)	Refractive Sphere (Mean Diopter ± SD)	Refractive Cylinder (Mean Diopter ± SD)	Visual Acuity UCVA (LogMar)	Visual Acuity BCVA (LogMar)
Hafezi F, 2007 (56)
10 P (10 e)	Baseline	57.4 D	NR	NR	NR	NR
36.2 years (R; 27 to 43)	12-month Postop	56.3 D 5/10 ↓ ≥ 2 D	7/10 ↓ ≥ 2 D	7/10 ↓ ≥ 2 D	NR	8/10 > 1 line
36.2 years (R; 27 to 43)	P-Value		NR	NR		NR
Salgado J, 2011 (57)
15 P (22 e)	Baseline	44.12 ± 3.97 D	−1.15 D	−2.59 D	0.53 ± 0.38	0.19 ± 0.21
38.4 years (R; 27 to 51)	12-month Postop	42.04 ± 2.67 D	−1.06 D	−2.10 D	0.40 ± 0.27	0.15 ± 0.14
38.4 years (R; 27 to 51)	P-value	NS	NS	NS	NS	NS
Vinciguerra P, 2010 (58)	Baseline	45.93 ± 6.03 D (R 37.42 to 57.01)	−2.96 ± 2.63 D (R −9.00 to 0.25)	−2.40 ± 2.06 D (R −6.00 to 0.0)	1.08 ± 0.43 (R 0.40 to 1.70)	0.16 ± 0.14 (R 0.00 to 0.40)
9 P (13 e) 42 years (R; 30 to 59)	12-Month Postop	42.49 ± 4.88 D (R, 35.49 to 49.12)	−2.25 ± 1.39 D (R, −5.00 to -0.50)	−2.00 ± 2.00 D (R, −5.00 to 0.0)	0.94 ± 0.46 (R 0.01 to 1.30)	0.06 ± 0.08 (R 0.00 to 0.22)
9 P (13 e) 42 years (R; 30 to 59)	P-Value	NS	NS	NS	NS	< 0 .05

Author, Year	Patients (eyes)	Study Design	Objective
Chan C, 2007 (60)	21 P (25 e)	Matched case control	To compare ICRS implant versus ICRS and simultaneous CXL
Coskunseven E, 2009 (61)	43 P (48 e)	RCT 2- arm	To compare 2 sequences at 7 month intervals: CXL + ICRS Vs ICRS + CXL
El-Raggal T, 2010 (62)	16 P (10 e)	RCT 2-arm	To compare sequential (ICRS and CXL 6 months postop) or simultaneous order of ICRS then CXL
El-Raggal T, 2011 (63)	11 P (14 e)	RCT 3-arm	To evaluate the effect of CXL on laser channel creation in subsequent ICRS 6 months postop
Ertan A, 2009 (64)	17 P (25 e)	Longitudinal cohort	To evaluate sequential transepithelial CXL 6 months after ICRS
Vicente L, 2010 (65)	10 P (14 e)	Longitudinal cohort	To evaluate simultaneous ICRS and transepithelial CXL

Author	Patients (eyes)	Study Design	Objective
Kanellopoulos A, 2007 (67)	1 P (1 e)	Case report	Effectiveness of sequential topography-guided photorefractive keratectomy with CXL in progressive KC
Kymionis G, 2009 (68)	1 P (2 e)	Case report	Effectiveness of simultaneous photorefractive keratectomy and CXL on progressive pellucid marginal corneal degeneration
Kymionis G, 2010 (70)	1 P (1 e)	Case report	Effectiveness of simultaneous transepithelial phototherapeutic keratectomy and CXL to improve visual outcome in progressive KC
Kymionis G, 2009 (69)	12 P (14 e)	Case series	Effectiveness of simultaneous topography-guided photorefractive keratectomy and CXL for stability and visiÓn improvement in KC
Kymionis G, 2010 (71)	23 P (28 e)	Case series	Effectiveness of simultaneous photorefractive keratectomy and CXL in progressive KC for stability and functional visiÓn
Kymionis G, 2010 (72)	2 P (4 e)	Case reports	Effectiveness of simultaneous conductive keratoplasty and CXL for correction of irregular astigmatism in advanced KC
Stojanovic A, 2010 (73)	12 P (12 e)	Case series	Effectiveness of simultaneous topography-guided transepithelial surface ablation and CXL in KC and pellucid marginal corneal degeneration
Kanellopoulos A, 2011 (74)	22 P (32 e)	Case series	Effectiveness of simultaneous topography-guided partial transepithelial photorefractive keratectomy and CXL in post-LASIK ectasia

Complication		Author Year	Case	Organism /agent	Outcome
1	Keratitis noninfectious	Angunawela R, 2009 (83)	40-year-old (NR)	Nonidentified (Sterile infiltrates with overlying epithelial ulceration and corneal melting)	Decreased visual acuity
2	Keratitis noninfectious	Mangioris G, 2010 (86)	25-year-old Female	Nonidentified (Multiple stromal infiltrates)	Decreased visual acuity
3	Keratitis noninfectious	Rodriquez-Ausin P, 2011 (76)	21-year-old (NR)	Nonidentified (Multiple stromal infiltrates)	Placement of Intraocular lens
4	Keratitis noninfectious	Rodriquez-Ausin P, 2011 (76)	11-year-old Male	Nonidentified (Multiple stromal infiltrates)	Decreased visual acuity
5	Diffuse conjunctivitis	Gokhale N, 2010 (84)	19-year-old Male	Nonidentified	Corneal melt/perforation) and corneal transplant
6	Keratitis noninfectious	Labiris G, 2011 (85)	23-year-old Male	Nonidentified	(Corneal melt/perforation) and corneal transplant
7	Keratitis noninfectious	Koppen C, 2009 (89)	28-year-old Female	Nonidentified (Strong ciliary flush and presence stromal white infiltrates)	Residual central and superior scarring with decreased visual acuity
8	Keratitis noninfectious	Koppen C, 2009 (89)	17-year-old Male	Nonidentified (Strong ciliary flush and presence of stromal white infiltrates)	Central scarring led to reduced visual acuity resulting in a corneal transplant
9	Keratitis noninfectious	Koppen C, 2009 (89)	23-year-old Male	Nonidentified (Strong ciliary flush and presence of stromal white infiltrates)	Responded to treatment and initial visual acuity returned
10	Keratitis noninfectious	Koppen C, 2009 (89)	31-year-old Male	Nonidentified (inflammation with corneal edema and infiltrates, irregular epithelium)	Decreased visual acuity
11	Viral keratitis	Kymionis G, 2007 (78)	21-year-old Female	Herpes simplex	corneal opacity
12	Bacterial keratitis	Perez-Santonja J, 2009 (79)	29-year-old Female	Staphylococcus epidermides	residual haze
13	Bacterial keratitis	Pollhammer M, 2009 (80)	42-year-old (NR)	Escherichia coli	Decreased visual acuity
14	Parasitic keratitis	Rama P, 2009 (81)	32-year-old Male	Acanthamoeba,	corneal melt and corneal transplant
15	Bacterial keratitis	Sharma N, 2010 (82)	19-year-old Female	Pseudomasaeruginosa	Planned keratoplasty
16	Polymicrobiol keratitis	Zamora K, 2009 (77)	32-year-old Male	Streptococcus salivarius, Streptococcus oralis, coagulase-negative Staphylococcus sp.	Residual central corneal stromal haze and subepithelial scar

Author, Year	Haze Occurrence	Finding	Outcome
Caporossi A, 2010 (41)	36% (16/44)	Temporary haze	Resolved with steroids
Doors M, 2009 (52)	0% (0/29)	No corneal haze detected
Greenstein S, 2010 (94)	44 patients	Haze dosimetry (Scheimpflug) and graded (slit lamp 0 to+4), haze greatest at 1 month, plateaued at 3 months, decreased between 3-and 12-months but still (p <.001) elevated at 12-months over baseline	Absolute haze degree associated with poorer vision
El-Raggal T, 2009 (53)	100% (9/9)	All eyes developed faint diffuse stromal haze	Resolved at 1 month
Koller T, 2009 (45)	100% (21/21)	All eyes showed stromal haze at 1 month with a demarcation line in deeper stroma in 18 patients	At 6 months all corneas were clear except in 2 cases with discrete scarring in deep stroma
Lim L, 2011 (95)	7% (2/30)	Permanent deep stromal scar near apical cone away from visual axis	Not affecting vision at this time
Mazzotta C, 2008 (96)	11% (5/44)	Transient corneal opacity similar to haze within first 3 months	Resolved with topical steroids
Vinciguerra P, 2009 (42)	14% (4/28)	Grade +1 haze (Hanna scale)	Regressed after 1 month with topical steroids

Author Year	Patients	Baseline CCT (µm) Mean ± SD	3 months CCT(µm) Mean ± SD	Baseline to 3 months CCT (µm)	12-months CCT (µm) Mean ± SD	Baseline to 12-months CCT (µm)	P-value Baseline to 12-months
Arbelaez M, 2009 (50)	19	464 ± 27	439 ± 43	−25	464 ± 37	0	> 0.05
Caporossi A, 2010 (41)	44	451	449	− 2	450	−1	> 0 .05
Doors M, 2009 (52)	29	495 ± 48 (R, 427 - 618)	NR	−28 ± 23	NR	−24 ± 19	0.02
Henriquez M, 2011 (44)	10	463	− 12	475	463	−12	0.03
Grewal D, 2009 (97)	102	459 ± 40	436 ± 41	−23	450 ± 51	− 9	0.65
Saffarian L, 2010 (54)	53	461 ± 47 (R, 400 − 576)	NR	NR	445 ± 42 (R, 395 − 570)	−16	< 0.05
Vinciguerra P, 2009 (42)	28	491 ± 31	NR	NR	471 ± 29	−20	< 0.05

Author, Year Patients	Observation Point	GAT-IOP Mean ± SD mm/HG	IOPcc - ORC Mean ± SD mm/HG	IOPg-ORC Mean ± SD mm/HG
Goldich Y, 2009 (101)	Baseline	13.6 ± 2.06	10.2 ± 1.63	10.1 ± 1.66
Goldich Y, 2009 (101)	1 week Postop	16.7 ± 2.40 (P = .01)	14.1 ± 2.21 (P < 0 .001)	14.2 ± 2.73 (P < 0.001)
10 P	1-month Postop	16.5 ± 3.57 (P=04)	13.2 ± 3.55 (P= 0.03)	13.5 ± 3.12 (P= 0.01)
	3-months Postop	15.3 ± 3.48 P=.19)	11.6 ± 3.17 (P= 0.023)	12.0 ± 1.25 (P= 0.01)
	6-months Postop	14.7 ± 2.87 (P=.21)	11.2 ± 2.89 (P= 0.33)	12.2 ± 1.14 (P= 0.16)
Sedaghat M, 2010 (103)	Baseline	ND	13.98 ± 2.9	10.47 ± 10.07
51 P	6-months Postop	ND	13.14 ± 2.8 (P= 0.027)	10.07 ± 3.0 (P= 0.281)
Caporossi A, 2010 (41)	Baseline	14.773 ± 1.696 (R, 11 to 18)	ND	ND
44 P	6-months Postop	14.795 (P>.05)
44 P	12-months Postop	14.932 (P>.05)
Doors M, 2009 (52)	Baseline	11.9 ± 3.5(R, 7 to 20)↑	ND	ND
29 P	1,3,6 and 12-months Postop	NS (values not reported)
Raiskup-Wolf, F, 2008 (6) 130 P	Baseline change at 12-months Postop	0.2 ± 1.4 (P>.05)	ND	ND
Wollensak G, 2003 (7)	Baseline	13.6 ± 2.0	ND	ND
22 P	Last follow-up	13.8 ± 2.5 (P = .612)
Kymionis G, 2010 (102) 55 P	Baseline	9.95 ± 3.01 (R, 5 to 19)	ND	ND
	6-months Postop	11.40 ± 2.89 (P<.001) (R, 7 to 19)
	12-months Postop	11.35 ± 3.38 (P<.001) (R, 6 to 25)

Author Year	Patients	Preop 2 cells/mm Mean ± SD	Postop 2 cells/mm Mean ± SD	Pre-Post Change	P-Value
Kymionis G, 2009 (99)	15	2780 ± 197	2713 ± 116	− 2.4% at 1 month	0.14
Caporossi A, 2010 (41)	44	2451 ± 130 (R, 2092 to 3016)	2444 R (nr)	−2.0% at 12-months	> 0.05
Doors M, 2009 (52)	29	2701 ± 352 (R, 2071 to 3803)	2703 ± 273	unchanged	> 0.05
Henriquez M, 2011 (44)	10	2566	2464 at 6 months, 2484 at 12-months	− 3.3% at 12-months	> 0.05
Leccisotti A, 2010 (46)	51	2765 ± 176	2792 ± 146	+ at 6-months	> 0.05
Vinciguerra P, 2009 (42)	28	2651 ± 321	2598 ± 564	−2.0% at 12-months	0.13
Wollensak G, 2003 (7)	22	NR	NR	unchanged	0.45
Vinciguerra P, 2009 (42)	9	2555 ± 470 R(1515 to 2994)	2120 ± 517 R(1547 to 2857)	− 17% at 12-months	> 0.05

Author, Year	Patients	Ocular Structure	Findings
Henriquez M, 2011 (44)	10	Macular thickness	By optical coherence tomography - decreased thickness compared to baseline (216 µm) at 1, 3, 6 and 12-months 191 µm, 191µm, 198 µm and 200 µm
Grewel D, 2009 (97)	102	Lens density	By Scheimpflug imaging - no change in lens density (8.7 µm to 8.7 µm]
		Retinal nerve fiber layer	By Scheimpflug imaging - no change in retinal nerve fiber layer (101 µm to 103 µm)
		Foveal thickness	By Scheimpflug imaging - decreased foveal thickness NS -change from baseline to 12-months 175.7 ± 35.6 µm to 146.4 ± 8.5 µm
Knappe S, 2011 (104)	8	Junction region between Bowman’s membrane and the sub-basal nerve plexus	By confocal microscopy the preoperatively tortuous and branched nerve fiber pattern of the sub-basal nerve plexus typical of KC could not be visualized in the central cornea immediately postoperatively and at 3-months postoperatively. At 4 months, the sub-basal nerve fibers again became visible on confocal microscopy.
Kymionis G, 2009 (90)	10	Subepithelial nerve plexus	By confocal microscopy– subepithelial nerve plexus absent at 1 month and regenerated between 3 to 6 months
Mazzotta C, 2006 (105)	10	Subepithelial stromal nerve fibers	By confocal microscopy – normal regeneration and morphological structure of corneal epithelium after 5 days, complete absence of nerve fibers 15 to 30 days post with regeneration continuing and restoring corneal sensitivity after 6 months. No evidence of altered corneal transparency.
Mazzotta C, 2008 (92)	40	Limbus	By confocal microscopy- no damage to the limbal region
		Epithelial nerve plexus	By confocal microscopy, disappearance of subepithelial plexus and anterior-midstromal nerve fibers, restored around 6 months
		Epithelial layer	By confocal microscopy, epithelial regrowth completed after 4 days

PERMALINK

Collagen Cross-Linking Using Riboflavin and Ultraviolet-A for Corneal Thinning Disorders

G Pron

L Ieraci

K Kaulback

Executive Summary

Objective

Subject of the Evidence-Based Analysis

Clinical Need: Target Population and Condition

Description of Technology/Therapy

Evidence-Based Analysis Methods

Inclusion Criteria

Exclusion Criteria

Summary of Evidence Findings

Considerations for Ontario Health System

Conclusion

Keywords

Background

Objective of Analysis

Clinical Need

Etiology

Diagnosis

Disease prevalence and natural history

Table 1: Baseline Visual Acuity of Keratoconus Patients in the CLEK Cohort Study*.

Disease Management

Table 2: Baseline Visual Correction in Patients in CLEK Cohort Study*.

Vision Quality of Life

Table 3: Patient-based Time Tradeoff Utility Values Associated with Visual Acuity Levels in the Better Seeing Eye.

Table 4: Patient-based Time Tradeoff Utility Values Across Disease States.

Surgical Interventions

Corneal Transplant

Corneal Collagen Cross-Linking

Regulatory Status

Treatment Procedures

Mechanism of Action

Potential Risks and Limitations

Clinical Indications

Table 5: Reports of Other Indications for Corneal Collagen Cross-Linking*.

Evidence-Based Analysis

Research Question(s)

Methods

Search Strategy

Databases Searched

Inclusion Criteria

Exclusion Criteria

Additional Information Sources

Assessment of Quality of Evidence

Results of Evidence-Based Analysis

Other Systematic Reviews

Medical Advisory Secretariat Systematic Evidence Review

Table 6: Evidence Base for CXL for Corneal Thinning Disorders*.

Table 7: Level of Evidence Summary for CXL Management of Corneal Thinning Disorders*.

Section A. Effectiveness of CXL for Keratoconus

Table 8: Topographic Outcomes in Longitudinal Pre-Post Cohorts after CXL in Treated Eyes*.

Table 9: Topographic Outcomes in Longitudinal Pre-Post Cohorts after CXL in Treated and Untreated Fellow-Eyes*.

Table 10: Impact of CXL on Higher Order Topographic Outcomes*.

Table 11: Impact of CXL on Higher Order Topographic Outcomes-Comparison of Treated and Untreated Fellow-Eye*.

Refraction

Table 12: Change in Refractive Sphere at 12-Months Following CXL for Keratoconus*.

Table 13: Change in Refractive Cylinder at 12-Months Following CXL for Keratoconus*.

Visual Acuity

Table 14: Impact of CXL on Best Corrected Visual Acuity at 12-Months*.

Treatment Failure

Table 15: Treatment Failure Assessed by Topographic Outcome Measures at 12-Months*.

Contact Lens Tolerance Following CXL

Vision-Related Quality of Life and Patient Satisfaction

Section B. Effectiveness for Post-Lasik Induced Corneal Thinning

Table 16: CXL Treatment of Keratectasia After Laser In-Situ Keratomileusis*.

Section C. Adjunct Interventions with CXL

CXL and Intrastromal Corneal Ring Segments

Table 17: Adjunct Procedures with CXL Involving Intrastromal Corneal Ring Segments.

CXL and Photorefractive Keratectomy

Table 18: Adjunct Procedures with CXL Involving Photorefractive Keratectomy*.

CXL and Intraocular Lens Implants

Section D. Safety

Infection

Table 19: Complication Case Reports Following CXL in Keratoconus*.

Stromal Keratocytes

Corneal Stromal Haze

Table 20: Corneal Stromal Haze after CXL.

Table 1: Baseline Visual Acuity of Keratoconus Patients in the CLEK Cohort Study^*.

Table 2: Baseline Visual Correction in Patients in CLEK Cohort Study^*.

Table 5: Reports of Other Indications for Corneal Collagen Cross-Linking^*.

Table 6: Evidence Base for CXL for Corneal Thinning Disorders^*.

Table 7: Level of Evidence Summary for CXL Management of Corneal Thinning Disorders^*.

Table 8: Topographic Outcomes in Longitudinal Pre-Post Cohorts after CXL in Treated Eyes^*.

Table 9: Topographic Outcomes in Longitudinal Pre-Post Cohorts after CXL in Treated and Untreated Fellow-Eyes^*.

Table 10: Impact of CXL on Higher Order Topographic Outcomes^*.

Table 11: Impact of CXL on Higher Order Topographic Outcomes-Comparison of Treated and Untreated Fellow-Eye^*.

Table 12: Change in Refractive Sphere at 12-Months Following CXL for Keratoconus^*.

Table 13: Change in Refractive Cylinder at 12-Months Following CXL for Keratoconus^*.

Table 14: Impact of CXL on Best Corrected Visual Acuity at 12-Months^*.

Table 15: Treatment Failure Assessed by Topographic Outcome Measures at 12-Months^*.

Table 16: CXL Treatment of Keratectasia After Laser In-Situ Keratomileusis^*.

Table 18: Adjunct Procedures with CXL Involving Photorefractive Keratectomy^*.

Table 19: Complication Case Reports Following CXL in Keratoconus^*.

Table 21: Pachymetry Measures of Central Corneal Thickness in Keratoconus Following CXL^*.

Table 22: Change in Intraocular Pressure Following CXL^*.

Table 23: Impact of CXL on Endothelial Cell Density^*.

Table 24: Impact of CXL on Critical Ocular Structures^*.

Table 25: GRADE Quality of Evidence for CXL of Keratoconus^*.

Table 27: Total Annual Costs of CXL for Keratoconus – Estimated Physician, Clinic and Medication Costs^*.

Table A1: Effectiveness Reports of Corneal Collagen Cross-Linking of Keratoconus^*.

Table A2: Adjunct Interventions with Corneal Collagen Cross-Linking and Intrastromal Corneal Ring Segments^*.

Table A3: Adjunct Interventions with Corneal Collagen Cross-Linking and Photorefractive Keratectomy^*.

Table A4: Safety Studies of Corneal Collagen Cross-Linking^*.

Table A5: Complication Case Reports for Corneal Collagen Cross-Linking^*.

Table A6: Impact of CXL on Corneal Curvature in Keratoconus – Pre-Post Longitudinal Higher Order Topographic Outcomes^*.

Table A7: Impact of CXL on Corneal Curvature in Keratoconus − Higher Order Topographic Findings in Treated and Untreated Fellow-Eyes^*.

Outcome	Design	Quality	Consistency	Directness Appropriate Range of Patients	Overall Quality
Safety	Case reports, observational longitudinal pre-post study, comparative pre-post studies (treated to untreated fellow-eye) and 1 RCT	Moderate	Low complication rates reported	Inception cohorts specified with appropriate range	Moderate
Corneal Topography	Observational longitudinal pre-post study, comparative pre-post studies (treated to untreated fellow-eye) and 1 RCT	Moderate	Clinically relevant and statistically significant improvements in treated eye and regression in untreated fellow-eyes	Inception cohorts specified with appropriate range	Moderate
Visual Acuity	Observational longitudinal pre-post study, comparative pre-post studies (treated to untreated fellow-eye) and 1 RCT	Moderate	Clinically relevant and statistically significant improvements, although less predictable than topography	Inception cohorts specified with appropriate range	Moderate
Durability	Observational longitudinal pre-post study	Low	Limited data beyond 2-year follow-up	Inception cohorts specified with appropriate range	Low

Clinical phase	Item	Cost
Preoperative
Ophthalmologist consultation	OHIP fee A235 - Consultation	$71.30
Ophthalmologist repeat visit	OHIP fee A234 - Partial assessment	$25.10
Corneal topography	First visit measurement	$50.00
	Second visit measurement	$50.00
	Total cost	$196.40
CXL Procedure
Medication	Proparacaine 0.5% - local anaesthesia *	$23.76
	Riboflavin 0.1% - every 2-3 min. for 30 min. *	$19.50
Technical	UV-X Illumination device (UVA irradiation) *	$122.32
	General surgical and technical (disposables, technician) *	$149.00
Ophthalmologist procedure	Professional fee - substitute OHIP fee E117 “limited” keratectomy *	$308.30
	Epithelial abrasion - substitute OHIP fee Z871 - epithelial debridement *	$26.60
	Pachymetry - substitute OHIP fee G813 - corneal pachymetry *	$12.75
	Total cost (1 eye)	$662.23
	Total cost (2 eyes)	$1,324.47
Postoperative
Ophthalmologist repeat visit	OHIP fee A234 - Partial assessment (First day)	$25.10
	OHIP fee A234 - Partial assessment (First week)	$25.10
	OHIP fee A234 - Partial assessment (First month)	$25.10
Corneal topography	First month measurement	$50.00
Medication	Antibiotic - moxifloxacin (Vigamox) *	$30.00
	Anti-inflammation (corticosteroid) - dexamethasone (Maxidex) *	$22.19
	Total cost (1 eye)	$177.49
	Total cost (2 eyes)	$229.68
Total aggregated costs
	Total cost per patient (1 eye)	$1,036.12
	Total cost per patient (2 eyes)	$1,750.55

Description	Existing cases (Prevalence)	New cases (Incidence)	Average annual
Number of patients
Number of CXL keratoconus cases	4,047	148	1,497
Total cost - 1 eye per patient
Physician costs	$2.71M	$0.10M	$1.00M
Clinic costs	$1.10M	$0.04M	$0.41M
Medication costs	$0.39M	$0.01M	$0.14M
Total cost	$4.19M	$0.15M	$1.55M
Total cost - 2 eyes per patient
Physician costs	$4.12M	$0.15M	$1.52M
Clinic costs	$2.20M	$0.08M	$0.81M
Medication costs	$0.77M	$0.03M	$0.29M
Total cost	$7.08M	$0.26M	$2.62M
Average total cost
Physician costs	$3.41M	$0.13M	$1.26M
Clinic costs	$1.65M	$0.06M	$0.61M
Medication costs	$0.58M	$0.02M	$0.21M
Total cost	$5.64M	$0.21 M	$2.09M

Author, Year	Site Country	Study Design and Follow-Up	Population	Study Objective
Agrawal V, 2009 (49)	Clear Vision Eye Center, India	Retrospective pre-post longitudinal consecutive cohort Mean 10.1 ± 3.55 month (R, 6 to 16 months)	25 P − 37 e Progressive KC Mean age 16.9 yrs± 3.5 (R, 12 to 39 yrs)	To assess the impact of CXL at 1-yr follow-up in an Indian cohort affect with progressive KC
Arbelaez M, 2009 (50)	Muscat Eye Center, Oman	Prospective pre-post longitudinal cohort 1 year	19 P (14 M, 5 F) − 20 e Progressive moderate to severe bilateral KC Mean age 24.4 yrs (R,18 to 44 yrs)	To evaluate safety and effectiveness of CXL in improving visual acuity and stabilizing progression of KC
Caporossi A, 2006 (8)	Dpt Ophthalmology Siena University, Italy	Pre-post longitudinal cohort with untreated fellow-eye as control 6 months	10 P (8M, 2F) − 10 e Bilateral progressive low or moderate KC Mean age 31.4 yrs(R, 21 to 39 yrs)	To assess the effectiveness of CXL in reducing KC progression and improving vision
Caporossi A, 2010 (41)	Dpt Ophthalmology Siena University, Italy	Prospective pre-post longitudinal cohort with untreated fellow-eye as control 4 year	44 P − 44 e Progressive KC Age range 10 to 40 yrs	To assess the long term results of CXL for progressive KC
Doors M, 2009 (52)	Dpt Ophthalmology University Medical Center, Netherlands	Pre-post longitudinal consecutive cohort Mean 6.3 months ± 3.7 (R, 1 to 12-months	29 P − 29 e 28 progressive KC, 1 post-LASIK ectasia Mean age 35.1 yrs± 11.7 (R, 19 to 76 yrs)	To investigate the stromal demarcation line after CXL with optical coherence tomography and its impact on short term results in progressive KC
El-Raggal T, 2009 (53)	Dpt Ophthalmology, Ain Shams University Egypt	Pre-post longitudinal cohort 6 months	9 P (3 M, 6 F) − 15 e KC (Krumeich grade 1 − 111) Mean age 26.4 yrs (R, 21 to 31)	To evaluate the safety and effectiveness of CXL in reducing KC progression and evaluate the visual and refractive changes
Hersch P, 2011 (47)	Cornea and Laser Eye Institute – Hersch Vision Group and Dpt Ophthalmology New Jersey Medical School, New Jersey	Multicenter prospective RCT 1 year	58 P − 71 e (49 KC, 22 post-LASIkectasia) sham control 41 e (28KC, 13 post-LASIK ectasia) and control group 30 e (21 KC, 9 post-LASIK ectasia)	To evaluate the 1-year outcomes of CXL for treatment of progressive KC and LASIK<or photorefractive keratectomy induced corneal ectasia
Henriquez M, 2011 (44)	Oftalmo Salud Institute de Ojos, Peru	Pre-post longitudinal cohort and comparative untreated progressive KC control group 1 year	10 P (8 M, 2 F) − 10 e Progressive KC (Krumeich grade 1, 11) Mean age 29.7 yrs (R, 15 to 43)	To evaluate safety and efficacy of CXL for the treatment of progressive KC
Koller T, 2009 (55)	Institute for Refractive and Ophthalmic Surgery, Switzerland	Pre-post longitudinal cohort with untreated fellow-eye as control 1 year	21 P (15M, 6 F) − 21 e Mild to moderate KC (n = 8), pellucid marginal degeneration (n = 4) mixed (n = 9)	To compare by Scheimpflug imaging changes in corneal geometric shape after CXL in CXL treated and untreated cases progressive actasia
Leccisotti A, 2010 (46)	School of Biomedical Sciences, University of Ulster, United Kingdom	Pre-post longitudinal cohort with untreated fellow-eye as control 1 year	64 P Progressive KC Mean age 26.9 ± 6.3 yrs (R,18 to 39)	To evaluate clinical effects of trans-epithelial CXL in progressive KC
Raiskup-Wolf F, 2008 (6)	Department Ophthalmology, CG Carus University Hospital, Germany	Prospective pre-post longitudinal cohort Mean 26.7 months± 16.2 (R, 12-months to 6 years)	130 P − 241 e Progressive KC Mean age 30.04 yrs ± 10.46	To evaluate the long term effects of CXL in progressive KC
Saffarian L, 2010 (54)	Navid Didegan Eye Center, Iran	Prospective pre-post longitudinal cohort 1 year	53 P (31 M, 22 F)) − 92 e Progressive KC Mean age 21.5 yrs± 3.4 (R,16 to 30)	To evaluate outcomes of CXL for progressive KC in Iranian patients at 1 year
Vinciguerra P, 2009 (58)	Department Ophthalmology, Instituto Clinico Humanitas, Italy	Prospective pre-post longitudinal cohort with untreated as control 1 year	28 P (20 M, 8 F) − 28 e Progressive KC (grade 111 AK stage) with fellow untreated eye (1-11 stage) as control Range 24 to 52 years	To evaluate 1 year refractive, topographic, tomographic, and aberrometric outcomes after CXL for progressive KC
Vinciguerra P, 2010 (43)	Department Ophthalmology, Instituto Clinico Humanitas, Milano, Italy	Prospective pre-post longitudinal cohort with untreated as control 2 year	28 P (20 M, 8 F) − 28 e Progressive KC (grade 111 AK stage) with fellow untreated eye (1-11 stage) as control Range 24 to 52 years	To evaluate intra-operative and 2 year refractive, topographic, tomographic, and aberrometric outcomes after CXL for progressive KC
Wollensak G, 2003 (7)	Department Ophthalmology, Technical University of Dresden, Germany	Prospective pre-post longitudinal cohort with untreated fellow-eye as control Mean 23.2 months ±12.9 (R, 3 to 47 months)	22 P (12M, 10F) − 23 e Moderate to advanced progressive KC with fellow untreated eye as control Mean age 31.7 yrs± 11.9 (R,13 to 58)	To evaluate the safety and effectiveness of CXL on the progression of KC

Author, Year	Site, Country	Study Design	Population	Duration
Coskunseven E, 2009 (61)	Dunya Eye Hospital, Turkey	Prospective RCT − Group 1 CXL followed by ICRS implantation and Group 2 ICRS implantation followed by CXL − treatment interval was 7 months	43 P (25 M, 18 F) − 48 e KC AC Grade 1 - 111	6-month
Chan C, 2007 (66)	Private practice, California, United States	Comparative series with matched groups: Group 1 with ICRS implantation only versus Group 2 with trans-epithelial CXL performed after ICRS implantation on the same day	21 P − 25 e	Group 1 = 102 ± 39 days Group 2 = 97 ± 38 days
El-Raggal T, 2011 (63)	Ain Shams University, Cairo, Egypt	Prospective comparative series with patients undergoing CXL followed by ICRS placement 6 months later and subdivided into 3 groups based on the power settings of the femtosecond laser used in ICRS placement − Group 11,5 mJ power setting, Group 2 1.6 mJ power setting and Group 3 1.7 mJ power setting. The control group, Group 4, did not have CXL.	20 P − 20 e Progressive grade 11-111 KC Age range 23 to 32	6-month
El-Raggal T, 2011 (62)	Ain Shams University, Cairo, Egypt	Prospective randomized comparative series with patients undergoing ICRS implanted with femtosecond laser followed by CXL performed in one session (7 e) or performed 6 months apart (9 e)	10 P (4 M, 6 F) − 16e Progressive mild to moderate KC Mean age 27.9 ± 4.8 yrs (R, 22 to 36)	12-months
Ertan A, 2009 (64)	Department of Cataract and Refractive Surgery, Kudret Eye Hospital, Turkey	Pre-post comparative cohort − ICRS implantation followed at 6 months with trans-epithelial CXL	17 P (10 M, 5 F) − 25 e Bilateral KC Mean age 25.14yrs (R,16 to 39)	First postop visit (after INTACS) was 3.98 months second postop visit (after CXL) was 2.67 months
Kamburoglu G, 2008 (121)	Kudret Eye Hospital, Turkey	Case report, ICRS implantation followed by trans-epithelial CXL the following day for postoperative LASIK ectasia	27-year-old male Post-LASIK ectasia	8 months
Vicente L, 2010 (65)	Boxer Wachler Vision Institute, California, United States	Prospective cohort, Simultaneous trans-epithelial CXL following ICRS implantation	10 P − 14 e KC Mean age 35 ± 13 yrs (R, 13 to 58)	3 year

Author, Year	Site, Country	Study Design	Population	Duration
Kanellopoulos A, 2007 (67)	Laser Vision Institute, Greece	Case report, CXL followed at 12-months by a topographically-guided PRK with untreated fellow- eye as control	26-year-old male Bilateral progressive KC	230-months
Kymionis G, 2009 (68)	Institute Vision and Optics, University of Crete Medical school, Greece	Case report, YAG laser PRK followed by CXL on the same day followed 15 days later by the procedures in the second eye	34-year-old female, bilateral KC (pellucid marginal degeneration)	12-months
Kymionis G, 2010 (70)	Institute Vision and Optics, University of Crete Medical school, Greece	Case report. Simultaneous excimer laser photorefractive keratectomy followed by CXL	24-year-old male Progressive KC	6-month
Kymionis G, 2009 (69)	Institute Vision and Optics, University of Crete Medical school, Greece	Case series, simultaneous customized topography-guided surface Nd: YAG laser ablation followed by CXL	12 p − 14 e Progressive KC Mean age 28 yrs (R, 20 to 39)	10.69 months (R, 3 to 16)
Kymionis G, 2010 (71)	Institute Vision and Optics, University of Crete Medical School, Greece	Case series, Simultaneous photorefractive keratectomy (PRK) followed by CXL	23 p − 28 e KC Mean age 30 yrs± 9.35 (R, 20 to 44)	6-month
Kymionis G, 2010 (72)	Institute Vision and Optics, University of Crete Medical School, Greece	Case reports, Simultaneous conductive keratoplasty (CK) followed within 24 hours by CXL	2 p Bilateral KC22-year-old male, 23-year-old male	6-month
Stojanovic A, 2010 (73)	Eye Dpt University Hospital of North Norway, Norway	Case series, Simultaneous topographically-guided excimer laser surface ablation followed by trans-epithelial CXL	12 p − 12 e 6 KC, 6 PMD Mean age 39.8 ± 11.9 yrs (R, 26 to 62)	12-months
Kanellopoulos A, 2011 (74)	Laser Vision Institute, Greece	Case series, Simultaneous topographically-guided partial PRK followed by CXL	22 p − 32 e Post Lasik ectasia Mean age 32 yrs (R, 23 to 66)	27-months

Author, Year	Site, Country	Study Design and Follow-Up	Study Population	Study Objective
Bakke E, 2009 (122)	Dpt Ophthalmology, Universities Ulleval, Oslo and Northern Norway	Pre-post consecutive prospective cohort 1 week	30 P (27 M, 3F) − 30 e Mean age 31.2 yrs± 9.5 (R, 18 to 53)	To compare the severity of postoperative pain and rate of penetration of riboflavin between eyes treated with CXL and having epithelial removal either mechanically of by excimer laser
Goldich Y, 2009 (101)	Dpt Ophthalmology, Assaf Harofeh Medical Center, Tel-Aviv, Israel	Prospective pre-post cohort 6 months	10 P (7 M, 3 F) − 10 e Mean age 26.5 yrs± 5.7 (R, 18 to 37)	To assess biomechanical corneal properties, corneal hysteresis (CH) and corneal resistance (CRF) and compensated intraocular pressure (IOP)
Greenstein S, 2010 (94)	Cornea and Laser Eye Institute – Hersch Vision Group, CLEI Center for Keratoconus, New Jersey, United States	Prospective RCT − group 1: CXL treated (and fellow untreated eye as control) Group 2: sham riboflavin only 1 year	44 P (KC 31, post-LASIK ectasia 19) − 50 e and control group 41 e (28 KC, 13 post-LASIK ectasia)	To determine the natural history of CXL associated corneal haze with Scheimpflug 3-D corneal density and slit lamp biomicroscopic analysis
Greenstein S, 2011 (100)	United States	Multicenter prospective RCT 1 year	65 P [54 P (KC), 28 (post LASIK)] − 82 e with untreated fellow-eye as control and sham treated control group 41 e (28KC, 13 post-LASIK ectasia) and fellow untreated eye control 39 e (25 KC, 14 post-LASIK ectasia)	To evaluate the natural course of corneal thickness changes following CXL
Grewal D, 2009 (97)	Grewal Eye Institute Chandigarh, India	Pre-post longitudinal comparative cohort 1 year	102 P (55 M, 47 F) Progressive KC Mean age 25.6 yrs± 4.5 (R, 18 to 31 yrs)	To evaluate changes in corneal parameters using Scheimpflug imaging post CXL and to examine the effect of UVA exposure on lens density using rotating Scheimpflug imaging and foveal thickness using optical coherence tomography
Knappe S, 2011 (104)	Dpt Ophthalmology, University of Rostock, Germany	Prospective pre-post longitudinal cohort	8 P (6 M, 2 F) − 8 e Progressive KC Mean age 33.6 yrs± 13.9	To evaluate the time course of corneal structural changes following CXL using confocal in vivo microscopy
Koller T, 2009 (55)	Institute for Refractive and Ophthalmo Surgery, Zurich, Switzerland	Pre-post comparative cohort 1 year	99 P (62 M, 37F) − 117 e Progressive mild to moderate keratectasia (KC, PMD)	To evaluate failure and complication rates in first postoperative year of CXL
Kymionis G, 2009 (90)	Institute of Vision and Optics, University of Crete, Greece	Prospective pre-post comparative cohort 1 year	10 P − 10 e Progressive KC (n = 5), post-Lasik ectasia (n = 5) Normal (n = 3), Normal post-LASIK (n = 3)	Compared structural changes using corneal confocal microscopy in post-LASIK ectasia and KC following CXL
Kymionis G, 2009 (99)	Institute of Vision and Optics, University of Crete, Greece	Prospective pre-post cohort 1 month	15 P (10 M, 5 F) − 19 e Mean age 26.9 yrs± 6,5 (R, 17 to 40)	To evaluate peri-operative pachymetric changes following CXL
Kymionis G, 2010 (102)	Institute of Vision and Optics, University of Crete, Greece	Prospective pre-post longitudinal cohort 1 year	55 P (55 e) Mean age 24.4 yrs±4.1 (R, 18 to 36)	To examine the effect of CXL on intraocular pressure (IOP) by Goldman applanation tonometry (GAT)
Lim L, 2011 (95)	Corneal Eye Dpt, Singapore National Eye Center, Singapore	Prospective pre-post cohort 3 month	30 P (2 case reports both age 23) Progressive KC	To examine safety and efficacy following CXL and review complications of deep stromal scarring post CXL in mild KC
Mazzotta C, 2006 (105)	Dpt Ophthalmological Sciences, Siena University, Italy	Prospective pre-post cohort 6 month	10 P Progressive KC	To assess corneal tissue modifications and regeneration of epithelium and sub-epithelial nerve plexus by HRT 11 confocal microscopy following CXL
Mazzotta C, 2007 (91)	Dpt Ophthalmology and Neurosurgery, University of Siena, Italy	Prospective pre-post longitudinal cohort study 6 month	10 P (8 M, 2F) − 10 e Progressive low to moderate KC	To assess the ultrastructural modifications after CXL with HRT confocal microscopy in patients with progressive KC
Mazzotta C, 2007 (96)	Dpt Ophthalmology and Neurosurgery, University of Siena, Italy	Pre-post comparison with untreated fellow-eye as comparison 6 month	39 P − 40 e Bilateral KC (35 KC in Krumeich stage 1-11, 5 KC in stage 111)	To review 2 cases of stromal haze during the second and third postop month resistant to topical steroids
Mazzotta C, 2008 (92)	Dpt Ophthalmology and Neurosurgery, University of Siena, Italy	Pre-post longitudinal cohort 3 years	39 P − 44 e	To assess early and late corneal micro-morphological modifications with HRT 11 confocal microscopy following CXL
Mencucci R, 2007 (123)	Dpt Oto-Neuro-Ophthalmological Surgical Sciences, Eye Clinic, Italy	Case series Intra-operative	6 P (3 M, 3 F) − 6 e Bilateral progressive KC	To assess possible corneal thermal damage during CXL using in vivo surface thermographic analysis
Renesto A, 2010 (124)	Vision Institute, Dpt Ophthalmology, Federal University of Sao Paulo, Brazil	Prospective RCT (CXL group vs. riboflavin only) 3 months	32 P (9 M, 23 F) − 42 e CXL group (19 e): Mean age 29 yrs R, 17 to 55), Control group (22 e): Mean age 31 yrs (R, 22 to 55)	Cytological examination of ocular surface changes following CXL
Sedaghat M, 2010 (103)	Eye Research Center, Mashhad University of Medical Sciences, Iran	Pre-post comparative study 6 months	51 P (31 M, 20 F) − 56 e	To compare the impact of CXL on 2 corneal biomechanical parameters, corneal hysteresis (CH) and corneal resistance factor (CRF) by waveform analysis
Seiler T, 2006 (125)	Institute for Refractive and Ophthalmic Surgery, Zurich, Switzerland	Pre-post comparative cohort 2 weeks	16 P Progressive KC Mean age 26.4 yrs (R,18 to 39)	To investigate the CXL induced corneal stromal demarcation line between treated and untreated corneal stroma

Author, Year	Indication	Complication Event	Outcome
Angunawela R, 2009 (83)	Case Report, 40-year-old patient with progressive KC treated with CXL.	Non-infective keratitis, treated with preservative-free levofloxacin and dexamethasone 0.1% followed by prednisone acetate 1% and topical agents	Complete resolution of infiltrates but with residual marginal corneal thinning
Gokhale N, 2010 (84)	Case report, 19-year-old male with progressive bilateral KC treated bilaterally with CXL	1 week postoperatively presented with redness, watering, pain, loss of vision in his right eye. On exam right eye diffuse conjunctivitis, central corneal melt with severe thinning and perforation with adjacent edema.	A temporary cyano-acrylate glue applied bandage contact lens was applied until a donor cornea was available. A therapeutic keratoplasty was performed 5 days later. Histologic examination of the corneal button revealed central area of stromal loss with perforation. Stains for fungus and bacteria were negative.
Hafezi F, 2008 (87)	Case report, 33-year-old female. Post LASIK developed iatrogenic bilateral keratectasia during first pregnancy treated with CXL and resulting in exacerbation of keratectasia during second pregnancy	Bilateral iatrogenic keratectasia exacerbated during pregnancy	Not reported
Koppen C, 2009 (89)	Case report, 28-year-old female with bilateral progressive KC underwent CXL in right eye	On second postop day presented with redness, increasing pain and milky vision. On exam strong ciliary flush and presence of white superficial infiltrates. Cultures were negative and presumptive diagnosis of sterile keratitis that responded to high dose of topical corticosteroids.	At 1 year there was central and superior scarring in superficial stroma leading to a decreased visual acuity.
Koppen C, 2009 (89)	Case report, 17-year-old male with bilateral rapidly progressing KC underwent CXL in the left eye	On the second postop day presented with strong ciliary flush, white superficial infiltrates, and iritis with keratic precipitates. Cultures were negative. The epithelium defect healed slowly over 2 weeks. Iritis and keratitis responded to topical steroids.	At 7 months scar formation in the central stroma of the central cornea persisted leading to a decreased visual acuity resulting in subsequent keratoplasty.
Koppen C, 2009 (89)	Case report, 23-year-old male with KC and intolerant to RGP lenses underwent CXL in right eye	On fourth postop day, strong ciliary flush, white superficial infiltrates over the treated zone. Responded to treatment involving sub-conjunctival injection of steroids and changed antibiotics to ofloxacin.	At 6 months visual acuity returned to baseline. The other eye did not undergo CXL.
Koppen C, 2009 (89)	Case report, 31-year-old male with bilateral KC underwent successful CXL in the right eye 3 months previous to the CXL in the left eye	At the first postop day, signs of pronounced inflammation, corneal edema, irregular epithelium and corneal infiltrates. Treated with sub-conjunctival injection of steroids.	At 5 months post op, visual acuity was reduced over baseline
Kymionis G, 2007 (78)	Case report, 21-year-old female with bilateral KC treated by CXL in right eye and planned penetrating keratoplasty in left eye	Herpetic keratitis with iritis On day 5 geographic epithelial defect, stromal edema and cells in the anterior chamber treated initially with topical steroids then to acyclovir with diagnosis herpes simplex virus	Decrease in stromal edema, presence anterior chamber cells and at 2 months a mild central corneal opacity remained
Kymionis G, 2007 (88)	Case report, 27-year-old male, CXL 4 years post LASIK for iatrogenic ectasia	Diffuse lamellar keratitis On first postoperative day inflammation with infiltrates covering the interface including the central cornea	After an intensive course every 2 hours of corticosteroid (dexamethasone 1%) inflammation responded rapidly and by 9^th day infiltrates had resolved
Labiris G, 2011 (85)	Case report, 23-year-old male with bilateral progressive KC underwent CXL in the left eye	During the first postoperative day developed intense photophobia, watering and a nonspecific ocular discomfort. Intensive examination for autoimmune and infectious diseases were all within normal limits. Repeated cultures from the cornea and contact lens were negative.	Cornea presented slow re-epithelialization and progressive thinning resulting in descemetocele and perforation in the second month and underwent successful penetrating keratoplasty. Concluded that nonspecific irreversible damage to keratocytes resulting in corneal melting had occurred.
Mangioris G, 2010 (86)	Case report. 25-year-old female with bilateral KC underwent CXL.	Presented 5 days post CXL with multiple deep infiltrates in periphery of cornea near limbus. Corneal scraping and secretions were negative for organisms	At 2 months some nebulae remained in the central cornea and visual acuity was reduced.
Perez-Santonja J, 2009 (79)	Case report, 29-year-old female with progressive bilateral KC. CXL in right eye for stage 1 KC and 1 month later ICRS implants in the left eye for stage 11 KC.	Microbial keratitis, photophobia and blurring in the CXL treated right eye, culture proven Staphylococcus epidermidis treated with topical ofloxacin 0.3% and tobramycin at 1-hr intervals.	Ocular inflammation and corneal infiltrates improved rapidly and at 5 months postop a mild haze was detected
Pollhammer M, 2009 (80)	Case report, 42-year-old patient with KC treated with CXL and postoperatively experiencing pain and progressive reduction in visual acuity	Stromal infiltrates and anterior chamber inflammation due to bacterial infection with Escherichia coli	Infection successfully treated after several weeks with tobramycin and cephazolin eye drops Resulted in avascularized corneal scar and permanent reduction of visual acuity
Rama P, 2009 (81)	Case report, 32-year-old man with bilateral KC treated with CXL in the left eye reported conjunctival redness and discharge 3 days postoperatively	Progressive corneal involvement with corneal opacification and despite intensive oral and topical antibiotics and steroids for infection with Acanthamoeba, persisting severe inflammation with corneal ectasia and subtotal de-epithelialization	Corneal ulceration with melting and cornea perforation on day 11 followed by penetrating keratoplasty. At 2 months the graft was clear with no sign of infection
Rodriguez-Ausin P, 2011 (76)	Case reports, 21-year-old patient with bilateral progressive grade 111 KC treated with CXL in the left eye and 9 months later CXL in the right eye. Referred for slight pain and low visual acuity in right eye 48 hours postoperatively	Corneal infiltrates with ulcer and despite 3 weeks intensive oral and topical antibiotic treatment (cultures were negative or bacteria or fungi) stromal opacities persisted at 3 weeks and at 2 months corneal leucomas and inferior thinning	At one year, KC stabilization was reported and a bilateral toric ICL implantation significantly improved bilateral visual acuity
Rodriguez-Ausin P, 2011 (76)	Case report, 11-year-old male with grade 3 to 4 KC treated initially with contact lenses to improve visual acuity and subsequently treated with CXL for bilateral progressive KC	After CXL 48 hours sterile corneal infiltrates (cultures negative for bacteria, fungi or parasites) and after intensive antibiotic and steroid treatment at 3 months detected in right eye stromal haze and stromal inflammatory infiltrates	At 12-month F-up the right eye showed progression, worsening of best corrected visual acuity with contact lens (1 line lost) and grade 2 para-central scarring
Sharma N, 2010 (82)	Case report, 19-year-old female with KC underwent CXL in the right eye and on fourth postop day presented with 3-day history of pain, redness and decreasing vision in the right eye	On exam corneal infection along with a corneal ulcer with central infiltrates with epithelial defect involving 90% of the corneal depth. Corneal scrapings were positive for Pseudomonas aeruginosa which responded to antibiotic treatment	At 2 months infiltrates decreased in size and a leucomatous corneal opacity remained with greatly reduced visual acuity. An optical keratoplasty is planned for visual rehabilitation.
Zamora K, 2009 (77)	Case report, 32-year-old male with KC underwent CXL in the left eye and presented on the third postop day with a 1-day history of red painful eye.	On exam conjunctival infection, severe keratitis with central corneal epithelial defect, ring of infiltrates and dense fibrin reaction throughout the anterior chamber. Cultures from the contact lens were positive for Streptococcus salivarius and S oralis. And corneal scrapings were positive for Staphylococcus sp.	At 2 months, residual central corneal stromal haze and a sub-epithelial scar in a ring-like configuration remained

Author, Year	Patients (P) Mean Age± SD	Keratometric Outcome	Baseline Diopters Mean ± SD	Diopters Mean ± SD	Difference Diopters Mean ± SD	P-Value
Agrawal V, 2009 (49)	25 P (37 e) 16.9 ±3.5 years	K-Max, K-Apex 12-Month(Keratron Scout, Optikon, Italy)	53.26 ± 5.93 D		−2.47 ± 3.89 D −2.73 ± 7.95 D K-Max summary Decreased (± 1.0D) 54% (20/37) Stable (± 0.5D) in 38%(14/37) Worse (>-1.0 D) 8% (3/37)	0.004 0.004
		Corneal wavefront surface aberrometry			Spherical and higher order corneal aberrations did not show significant changes Coma component (lower order aberrations) showed significant reduction	Coma 0.003
Arbelaez M, 2009 (50)	19 P (20 e) 24.4 (R,18 to 44) years	K- Average K- Apex at 12-months	49.93 ± 5.02 D 51.89 ± 7.99 D	48.57 ± 4.54 D 50.42 ± 8.09 D	1.36 D 1.40 D	0.004 0.01
		Corneal wavefront surface aberrometry			Spherical and higher order corneal aberrations did not show significant changes Absolute RMS and absolute coma were significantly reduced	0.041 0.026
Doors M, 2009 (52)	29 P (29 e) 35.1 ±11.7 years (R, 19 to 76)	Central K K-Max at 6 months Corneal topography (Eye map), Pentacam HR tomography (Oculus),			Mean central K and K-Max were not significantly changed	> 0.05
		Corneal aberrometry (IRX-3 Wavefront Aberrometer)	Coma-x −0.05 ± 0.38 µm Coma-y -0.19 ± 0.92 µm Spherical aberration 0.12 ± 0.27 µm	−0.19 ± 0.39 µm −1.10 ± 0.91 µm 0.07 ± 0.30 µm	Higher-order aberration values coma-x, coma-y and spherical aberration were not significantly changed	> 0.05
El-Raggal T, 2009 (53)	9 P (15 e) 26.4 (R, 21 to 31) years	K-average	47.77 ± 1.79 D (R, 44.8 to 50.1)	46.27 ± 1.66 D (R, 43.6 to 48.7)		<0.001
		K-Max At 6 months	52.17 ± 1.66 D (R, 49.2 to 54.1)	50.42 ± 1.57 D (R, 47.9 to 52.7)		<0.001
		Corneal topography (TMS-4 Tomey Inc)
Raiskup-Wolf F, 2008 (6)	130 P (142 e) 30 ± 10.5 years	K-Apex			−2.68± 7.61 D	<0 .01
		K-Max At 12-months			−1.46 ± 3.76 D	<0.01
					K-Apex decreased in 62% eyes, remained stable in 17% and increased in 21
Saffarian L, 2010 (54)	53 P (92 e) 21.5 ±3.4 (R,16 to 30) years	Sim-K Corneal topography by Orbscan 11 −	46.94 ± 2.37 D	46.0 ± 2.33 D	0.94 ± 0.71 D	< 0.001

Author Year	Patients (P) Mean Age	Keratometric (K-Value) Outcome	Treated Eye Diopters Mean ± SD	Untreated Eye Diopters Mean ± SD	Pre-post change Treated Vs Untreated Eye
Caporossi A, 2010 (41)	44 P (44 e) Range 10 to 40 years	K- Average at 12-monthsCorneal topography (CSO Eye Top,	Pre-post change −1.96 ± 0.63 D (R, 0.92 to −3.24)	Pre-post change +1.2 ± 0.96
		Surface aberrometry (CSO Eye Top)	Pre-post change Total wavefront higher order aberrations significantly reduced (P<0.00001)
		Corneal symmetry- Inferior-Superior-inferior index (SI)	Pre-post SI asymmetry index significantly improved (P< 0.0001) Spherical aberration remained unchanged
Henriquez M, 2011 (44)	10 P (10 e) and 10 control P (10 e) 29.7 (R, 15 to 43) years	K-Max K-Min At 12-months Corneal typography Topcon	Pre-post reduction K-Max 2.66 D p =.04 K-Min 1.61 D (P= 0.03) 80% (8/10) decreased K-Max 20% (2/10) worsening K-Max	90% (9/10) had increased K-Max
Koller T, 2009 (45)	21 P (21 e) Age NR	Corneal topography Pentacam system with Scheimpflug camera − progression regression minimum corneal curvature radius (Rmin/mm) At 12-months	Pre-post improvement Rmim/mm 6.14 to 6.21 (P= 0.01)	Pre-post worsening Rmin/mm 6.94 to 6.86 (P= 0.002)	Pre-post change Rmin/mm 0.066 ± 0.10 vs −0.08 ± 0.10 (P = 0.0009) Rmin/D
			Rmin/D 55.0 D to 54.3D (P= 0.01)	Rmin/D 48.6 to 49.2 (P= 0.002)	−-0.62 ± 0.9 vs 0.57 ± 0.8 (P= 0.0009)
			4 / 7 Pentacam KC indices showed significant improvements towards a more normalized corneal anterior surface	Based on ARmin 0.12 mm ≃ 1 Progressed = 7 untreated Vs 0 treated Unchanged = 14 untreated Vs 13 treated Regressed = 0 untreated Vs 8 treated
Leccisotti A, 2010 (46)	51 P (51 e) 26.9 ± 6.3 (R 26.9 to 39) years	K-Apex Average Sim-K (averaging flattest and steepest meridian) At 12-months	Pre-post K-Apex 54.31 ± 8.76 to 54.81 ± 4.86 Mean Δ 0.51 ± 7.79 (P>0.05) Pre-post Sim-K 46.63 ± 2.89 to 46.53 ± 3.18 Mean Δ-0.10 ± 1.44 (P> 0.05)	Pre-post K-Apex 51.69 ± 6.42 to 53.30 ± 7.71 Mean Δ 1.61 ± 6.28 (P> 0.05) Pre-post Sim-K 44.60 ± 2.19 to 45.48 ± 2.88 Mean Δ 0.88 ± 2.35 (P> 0.05)	Comparative Pre-post Δ K-Apex (95% CI) −3.98 to 1.87 Comparative Pre-post Δ Sim-K (95% CI) −1.75 to −0.21
		Corneal surface regularity	Pre-post	Pre-post	Comparative Pre-post Δ
		index (ISV) -deviation of individual corneal radii from mean values (abnormal >37) At 12-months Tangential video keratography (Keratograph	ISV 76.4 ± 28.2 to 77.3 ± 28.8 Mean Δ 0.9 ± 4.69 (P> 0.05)	ISV 48.2 ± 14.3 to 53.5 ± 19.5 Mean Δ 5.3 ± 7.30 (P> 0.05)	ISV (95% CI) 1.99 to 6.81
Vinciguerra P, 2009 (48)	28 P (28 e) Range 24 to 52 years	K-Min K-Max Sim-K	Pre-post K-Min 46.10 to 40.22 (P= 0.0003)
		Corneal topography with CSO Eye Top Topographer and 21-Klyce indices	K-Max 50.37 to 44.21 (P= 0.0011) Sim-K 48.08 to 42.01 (P= 0.0004)
		Total (corneal and internal) wavefront analysis performed with Nidek OPD-Scan - 21-Klyce Indices	Pre-Post Sim-K1 50.53 to 50.13 (P < 0.05)	Pre-Post Sim-K1 45.81 to 46.21 (P< 0.05)
			Sim-K2 45.89 to 45.76 (P < 0.05)	Sim-K2 43.29 to 43.57 (P < 0.05)
			K-Min 41.18 to 41.32 (P > 0.05)	K-Min 40.83 to 40.92 (P > 0.05)
Wollensak G, 2003 (7)	22 P (23 e) 31.7 ±11.9 (R 13 to 58) years	K-max At 12-months Videokeratoscope (C-scan)	Pre-post K-Max Mean regression of 2.01 ± 1.74 D (p =.03)
			K-Max Regression in 70% (16/23) with average reduction 2.01 D (95% CI; 1.23 to 3.07)	K-Max progression on average 1.48 D in 22% (5/22)
			Stable in 22% (5/23) Increase in 1 patient of 0.28 D
Hersh P, 2011 (47)	58 p (71 e) 41 KC and e	K-Max K-Average K-Min (Flat) K-Max (Steep) At 12-months Topography Scheimpflug camera (Pentacam)	Pre-post Treated Eye K-Max 60.4 ± 9.99 to 58.4 ± 8.41 (p <.05) K-Max Mean Δ 1.5 D (p<.001) K-Average 50.4 ± 7.06 to 48.9 ± 5.48 (p<.05) K-Steep 52.9 ± 7.45 to 51.5 ± 5.94 (p<.05) Δ K-Max Decrease > 2 D(N=22), 1−2 D (N = 14), No change (N=28), Increase 1−2 D (N = 4), lncrease>2 D	Pre-post Sham Group K-Max, K-Average, K-Steep - No statistically significant changes Pre-post Fellow-eye Δ K-Max +0.29 ± 1.19 D (p =.19) Δ K-average +0.20 ± 0.79 (p =. 18)	Comparative pre-post Δ Treatment Vs Sham Group Only 3 month comparison as all crossed over No significant differences in kerotometry Comparative pre-post Δ Treatment Vs Untreated Fellow-eye Δ K-Max p <.001 Δ K-Average p <.001