ORIGINAL PAPER

Corneal collagen cross-linking using riboflavinand ultraviolet-A irradiation: a review of clinicaland experimental studies

Maria Gkika • Georgios Labiris •

Vassilios Kozobolis

Received: 4 November 2010 / Accepted: 8 July 2011 / Published online: 17 August 2011

� Springer Science+Business Media B.V. 2011

Abstract Corneal collagen cross-linking (CXL)

using riboflavin and ultraviolet-A irradiation is a

common method of tissue stabilization and has been

developed primarily to address the need of treating

keratoconus. CXL’s promising results on keratoconus

indicated that it might be effective in other corneal

diseases as well. This new treatment promises a

slowing effect on the progression of these diseases

and its initial results show that it is safe and

reasonably curative. The purpose of this review is

to critically evaluate this treatment, to explore its

benefits, to highlight its limitations in terms of

efficacy and long-term safety and finally to identify

areas for future research in this topic with a

significant potential to change the way we treat our

patients. In addition, in this unbiased review we try to

bring together all the scientific information from both

laboratory and clinical trials that have been con-

ducted during recent years and to review the most

recent publications regarding the therapeutic indica-

tions of CXL.

Keywords Corneal collagen cross-linking �Riboflavin � Ultraviolet-A irradiation � Clinical and

experimental studies � Review article

Introduction

Cross-linking is a common method of tissue stabil-

ization. For example, cross-linking is used for

formaldehyde-induced tissue stiffening and fixation

in pathologic specimens. Corneal collagen cross-

linking (CXL) is performed with ultraviolet-A (UVA)

irradiation at 370 nm and the photosensitizer ribofla-

vin (vitamin B2). According to Wollensak et al. [1],

the photosensitizer is excited into its triplet state

generating reactive oxygen species (ROS), which are

mainly singlet oxygen and to a much lesser degree

superoxide anion radicals. ROS can react further with

various molecules inducing chemical covalent bonds

that form bridges between amino groups of collagen

fibrils (type II photochemical reaction). The biome-

chanical effect occurs immediately after irradiation

leading to an increase of the biomechanical rigidity of

the cornea of about 300%. The optimal wavelength of

the ultraviolet (UV) radiation is 370 nm, at which

riboflavin presents maximal absorption. Interestingly,

a similar mechanism has been detected in the human

crystalline lens [2].

All patients who are candidates for CXL treatment

need to have maximum keratometric (K) readings

M. Gkika (&) � G. Labiris � V. Kozobolis

Eye Institute of Thrace and Department

of Ophthalmology, Medical School of Democritus

University of Thrace, Dragana, 68100 Alexandroupolis,

Greece

e-mail: mggkika@yahoo.gr

123

Int Ophthalmol (2011) 31:309–319

DOI 10.1007/s10792-011-9460-x

less than 60 diopters (D) and a central corneal

thickness (CCT) of at least 400 lm. All clinical trials

of CXL refer to patients aged between 18 and

30 years. There are as yet no extensive clinical trials

in children and this is the reason why great care must

be taken when applying the CXL technique in

patients younger than 18. In addition, patients with

a history of herpetic keratitis, corneal scarring, severe

eye dryness, pregnancy or nursing, previous anterior-

segment surgery, systemic collagen pathology or

concomitant autoimmune diseases should be handled

with great caution and perhaps better excluded.

In brief, the CXL procedure is conducted under

sterile conditions in the operating room, after the

patient’s eye is anesthetized. According to the

Dresden therapeutic protocol, the central 8 mm of

the corneal epithelium are removed to allow better

diffusion of riboflavin into the stroma. Without

epithelial removal, the biomechanical effect is less

than 50% of the standard cross-linking procedure. In

fact, Bottos et al. [3] conclude in their study that

treatment of the cornea with riboflavin and UVA

without previous de-epithelialization did not induce

any cross-linking effect. Consequently, to facilitate

diffusion of riboflavin throughout the corneal stroma,

the epithelium should be removed as an important

initial step in the treatment [3]. Partial grid-pattern

epithelial removal allows some riboflavin penetra-

tion, but uptake is limited and non-homogeneous,

which may affect the efficacy of the cross-linking

process [4]. Following de-epithelialization, a 0.1%

riboflavin solution (10 mg riboflavin-5-phosphate in

10 ml dextran 20% solution) is instilled to the cornea

for 30 min (2 drops every 2 min) prior to the

irradiation, until the stroma is completely penetrated

and the aqueous humour is stained yellow. In human

and porcine corneas, riboflavin does not appear to

fully load the corneal stroma using the current

clinical procedure. Instead, the uptake appears to be

limited to the anterior approximately 200 lm.

Changes in application time and riboflavin concen-

tration have only little influence on stromal depth

diffusion [5]. Generally, the riboflavin film is an

integral part of the CXL procedure and important in

achieving the correct stromal and endothelial UVA

irradiance [6]. The irradiation is performed from

1 cm distance for 30 min using a UVA double diode

at 370 nm and an irradiance of 3 mW/cm2 (equal to a

dose of 5.4 J/cm2). The required irradiance is

controlled in each patient directly before the treat-

ment to avoid a potentially dangerous UVA overdose

[7, 8]. Instillation of riboflavin drops (1 drop every

2 min) is continued during irradiation as well, in

order to sustain the necessary concentration of

riboflavin. Moreover, balanced salt solution (BSS)

is applied every 6 min to moisten the cornea.

A series of variations of the treatment protocol

have been demonstrated. Boxer-Wachler et al. [9]

suggested CXL without de-epithelialization, while

Kanelopoulos et al. [10] demonstrated that CXL by

means of a femtosecond laser facilitated intrastromal

0.1% riboflavin administration, with promising pre-

liminary results. A limited but favourable effect of

trans-epithelial CXL was noted on keratoconic eyes,

without complications by Leccisotti et al. [11]. The

effect appears to be less pronounced than described in

the literature after CXL with de-epithelialization. In

addition Bakke et al. [12] attempted to compare the

severity of postoperative pain and the rate of

penetration of riboflavin between eyes treated by

CXL using excimer laser superficial epithelial

removal and mechanical full-thickness epithelial

removal, and concluded that superficial epithelial

removal using the excimer laser resulted in more

postoperative pain and the need for prolonged

application of riboflavin to achieve corneal

saturation.

The objective of this article is to review the most

recent publications regarding the therapeutic indica-

tions of CXL.

Laboratory studies

A series of laboratory studies attempted to explore all

aspects of CXL procedure and particularly its

biomechanical, thermomechanical, and morphologi-

cal impact on corneal cells. Microcomputer-con-

trolled biomaterial testing experiments indicated an

impressive increase in corneal rigidity of 71.9% in

porcine and 328.9% in human corneas, and an

increase in Young’s modulus by a factor of 1.8 in

porcine and 4.5 in human corneas after CXL.

Moreover, clinical observations suggest that ribofla-

vin/UVA-induced collagen cross-linking leads to an

increase in biomechanical rigidity which remains

stable over time [13, 14]. However, the cross-linking

effect was maximal only in the anterior 300 lm. The

310 Int Ophthalmol (2011) 31:309–319

123

greater biomechanical effect in human corneas was

attributed to the relatively larger portion of cross-

linked stroma because of the lower total corneal

thickness of 550 lm in human corneas compared

with 850 lm in porcine corneas [15].

In thermomechanical experiments with porcine

corneas, the maximal hydrothermal shrinkage tem-

perature was found to be 708�C for the untreated

controls, 758�C for the corneas cross-linked with

riboflavin and UVA and 908�C for corneas cross-

linked with glutaraldehyde, suggesting the correlation

between shrinkage temperature and CXL. The heat-

dependent denaturation of non-cross-linked collagen

could be demonstrated by the loss of birefringence in

histological sections [16].

In the anterior stroma of rabbit corneas treated

with CXL, the collagen fibre diameter was signifi-

cantly increased by 12.2% (3.96 nm), but by only

4.6% (1.63 nm) in the posterior stroma [17]. Similar

changes have been reported in the cornea and other

tissues due to age-related or diabetes mellitus-related

collagen cross-linking. A possible explanation for

this observation is that the induced cross-links push

the collagen polypeptide chains apart, resulting in

increased intermolecular spacing. On scanning elec-

tron microscopy, the collagen fibres of cross-linked

vitreous humour also appeared markedly thickened

and coarse [18].

Experiments in rabbits treated with riboflavin and

variable UVA irradiances ranging from 0.75 to

4 mW/cm2 indicated a variable apoptosis that was

proportional to the irradiance power. The cytotoxic

UVA irradiance level for keratocytes was determined

to be about 0.5 mW/cm2 [19]. The latter was also

confirmed by in vitro studies on keratocyte cell

cultures [20]. Cytotoxic apoptosis is followed by

repopulation approximately after 4–6 weeks from the

irradiation [21].

A series of observations suggest that the CXL

effect is maximal on the anterior part of cornea,

including: (a) the collagen fibre diameter is signifi-

cantly increased only in the anterior half of the

stroma, (b) in thermomechanical experiments with a

water bath temperature of 708�C a mushroom-shaped

appearance of the corneal astigmatism was demon-

strated due to the tissue shrinkage of the non-

crosslinked posterior stromal collagen [16], (c) in

enucleated porcine eyes CXL led to a significant

change in the swelling behaviour of the anterior

stroma [22], (d) significant biomechanical and bio-

chemical differences between the anterior and pos-

terior parts were demonstrated in post-CXL corneas

[23, 24]. Also, in a patient with acute keratoconus

(KC) who underwent CXL, a rapid progress of KC

developed 2 years after treatment. This was probably

due to the fact that CXL does not effectively treat the

posterior corneal layers and Descemet’s membrane,

which are mainly affected by acute KC [25]. Also, in

normal corneas, the anterior stroma is more rigid

because it is designed to maintain the anterior corneal

curvature. Interestingly, the anterior stroma of these

corneas after CXL is found to be more cross-linked

and still more rigid than the posterior one. This

degree of rigidity is even preserved in the presence of

corneal oedema [26].

In cross-linked porcine eyes, a markedly increased

resistance to collagenase digestion was described,

with a 15 day digestion time in the cross-linked

samples compared with 6 days in the controls [27].

This effect is stronger in the anterior half of the

cornea.

Corneal collagen cross-linking in keratoconus

Keratoconus is a relatively common disease with an

incidence of 1 in 2000 in the general population.

Current technology provides no effective means for

interrupting its progression, resulting to a disappoint-

ing 21% of KC patients that require corneal trans-

plantation. Prevalent therapeutic interventions, such

as intracorneal rings, photorefractive keratectomy

(PRK), and contact lenses attempt to improve the

refractive errors and not the pathomechanism itself.

On the other hand, CXL targets the biomechanical

properties of corneal collagen [7, 8, 28, 29]. The first

clinical study on the cross-linking treatment of KC

was performed by Wollensak et al. [7]. In this 3 year

study, 22 patients with progressive KC were treated

with riboflavin and UVA. In all treated eyes, the

progression of KC was interrupted. In 16 eyes, a

slight reversal and flattening of the KC by 2 D was

detected. Best corrected visual acuity (BCVA)

improved slightly in 15 eyes [7].

A series of studies confirmed Wollensak et al.’s

results and CXL’s beneficial impact on KC [27–29]

(Table 1). Sandner et al. [30] showed in their 5 year

follow-up study that none of the 60 KC eyes that

Int Ophthalmol (2011) 31:309–319 311

123

Ta

ble

1R

efer

ence

so

nco

rnea

lco

llag

encr

oss

-lin

kin

gas

atr

eatm

ent

for

ker

ato

con

us

and

ker

atec

tasi

ap

ub

lish

edin

the

last

5y

ears

wit

hlo

ng

erfo

llo

w-u

pan

da

larg

ern

um

ber

of

case

s

Ref

eren

ces

Nu

mb

er

of

case

s

Fo

llo

w-u

pK

val

ue

po

st-o

per

ativ

e

reg

ress

ion

BC

VA

po

st-o

per

ativ

e

reg

ress

ion

Ref

ract

ive

erro

r

po

st-o

per

ativ

e

reg

ress

ion

(D)

Ker

ato

con

us

Wo

llen

sak

etal

.[1

]

Rib

ofl

avin

/ult

rav

iole

t-A

-in

du

ced

coll

agen

cro

ssli

nk

ing

for

the

trea

tmen

t

of

ker

ato

con

us

Am

JO

ph

thal

mo

l1

35

:62

0–

62

7

23

eyes

Mea

n2

3.2

mo

nth

s

Ran

ge

3–

48

mo

nth

s

Mea

n2

.0D

Mea

n1

.3li

nes

Mea

n1

.1

Rai

sku

p-W

olf

etal

.[3

3]

Co

llag

encr

oss

lin

kin

gw

ith

rib

ofl

avin

and

ult

rav

iole

t-A

lig

ht

ink

erat

oco

nu

s:

Lo

ng

-ter

mre

sult

s

JC

atar

act

Ref

ract

Su

rg3

4(5

):7

96

–8

01

24

1ey

esM

ean

26

.7m

on

ths

Ran

ge

12

–7

2m

on

ths

Mea

n2

.3D

C1

lin

eM

ean

1.2

Ho

yer

etal

.[3

6]

Co

llag

encr

oss

-lin

kin

gw

ith

rib

ofl

avin

and

UV

Ali

gh

tin

ker

ato

con

us.

Res

ult

s

fro

mD

resd

en

Op

hth

alm

olo

ge

10

6(2

):1

33

–1

40

15

3ey

esM

ean

27

.9m

on

ths

Ran

ge

12

–9

0m

on

ths

Mea

n2

.2D

C1

lin

eM

ean

1.3

Vin

cig

uer

raet

al.

[34]

Intr

aop

erat

ive

and

po

sto

per

ativ

eef

fect

s

of

corn

eal

coll

agen

cro

ss-l

ink

ing

on

pro

gre

ssiv

ek

erat

oco

nu

s

Arc

hO

ph

thal

mo

l1

27

(10

):1

25

8–

12

65

28

eyes

Mea

n2

4m

on

ths

Ran

ge

–

Mea

n1

.1D

Mea

n2

.0li

nes

Mea

n0

.8

Gre

wal

etal

.[3

8]

Co

rnea

lco

llag

encr

oss

-lin

kin

gu

sin

g

rib

ofl

avin

and

ult

rav

iole

t-a

lig

ht

for

ker

ato

con

us:

on

e-y

ear

anal

ysi

su

sin

g

Sch

eim

pfl

ug

imag

ing

JC

atar

act

Ref

ract

Su

rg3

5(3

):4

25

–4

32

10

2ey

esM

ean

12

mo

nth

s

Ran

ge

–

Mea

n0

.9D

NS

1.4

Cap

oro

ssi

etal

.[3

5]

Lo

ng

-ter

mre

sult

so

fri

bo

flav

in

ult

rav

iole

ta

corn

eal

coll

agen

cro

ss-

lin

kin

gfo

rk

erat

oco

nu

sin

Ital

y:

the

Sie

na

eye

cro

ss-s

tud

y

Am

JO

ph

thal

mo

l1

49

(4):

58

5-5

93

44

eyes

Mea

n5

2.4

mo

nth

s

Ran

ge

48

–6

0m

on

ths

Mea

n2

.0D

Mea

n1

.9li

nes

Mea

n2

.1

312 Int Ophthalmol (2011) 31:309–319

123

were treated with CXL presented further keratectasia.

Moreover, 31 eyes presented an average post-CXL

regression of 2.87 D, while BCVA improved slightly

by 1.4 lines. In 2006, Caporossi et al. [31] in their

prospective non-randomized study attempted an abb-

erometric analysis that suggested an improvement in

morphological symmetry with a significant reduction

in chromatic aberrations. The aforementioned results

were confirmed by Witting-Silva et al. [32], who

conducted a randomized controlled trial of corneal

collagen cross-linking in 66 eyes of 49 patients with

progressive KC. Long-term corneal stabilization after

CXL was indicated by Raiskup-Wolf et al. [33], in

their 3 year follow-up study. Vinciguerra et al. [34]

reported that, 2 years postoperatively, corneal colla-

gen cross-linking appears to be effective in improving

uncorrected and best spectacle-corrected visual acu-

ities in eyes with progressive KC by significantly

reducing corneal average pupillary power, apical

keratometry, and total corneal wavefront aberrations.

Similar results were demonstrated by Caporossi et al.

[35] in the Siena eye cross-study and Hoyer et al. [36]

in their study, while Doors et al. [37] suggested that

anterior-segment optical coherence tomography (AS-

OCT) is a useful device for detecting the stromal

demarcation line after CXL. Also, in the same study,

their results showed that at 3–12 months follow-up,

all clinical parameters remained stable, indicating

stabilization of the keratoconic disease.

Regarding corneal ectasia, the literature suggests

that its regression progresses to some extent in about

50% of patients in the years after CXL. This is,

however, mostly a reduction by only 2–3 D and might

be attributed to the wound healing response. One-year

analysis of CXL treatment using Scheimpflug imaging

[38] indicated stable BCVA, spherical equivalent (SE),

anterior and posterior corneal curvatures and corneal

elevation, suggesting non-progressing KC. Vinciquer-

ra et al. [39] in their refractive, topographic, tomo-

graphic and aberrometric analysis of 28 KC eyes

undergoing CXL demonstrated improved mean BCVA

and uncorrected visual acuity (UCVA), while mean

SE, mean simulated keratometry flattest and steepest

meridians, SIM K average, mean average pupillary

power and apical keratometry were decreased. More-

over, contrary to control non-CXL KC eyes, CXL KC

eyes presented no deterioration of the Klyce indices.

Similar results were presented by Coskunseven et al.

[40] in a contralateral study of 38 eyes.Ta

ble

1co

nti

nu

ed

Ref

eren

ces

Nu

mb

er

of

case

s

Fo

llo

w-u

pK

val

ue

po

st-o

per

ativ

e

reg

ress

ion

BC

VA

po

st-o

per

ativ

e

reg

ress

ion

Ref

ract

ive

erro

r

po

st-o

per

ativ

e

reg

ress

ion

(D)

Ker

atec

tasi

a

Haf

ezi

etal

.[4

6]

Co

rnea

lco

llag

encr

oss

lin

kin

gw

ith

rib

ofl

avin

and

ult

rav

iole

t-A

totr

eat

ind

uce

dk

erat

ecta

sia

afte

rL

AS

IK

JC

atar

act

Ref

ract

Su

rg

33

(12

):2

03

5–

20

40

10

eyes

Up

to2

5m

on

ths

Mea

n1

.9D

Mea

n3

lin

esM

ean

1.2

Vin

cig

uer

raet

al.

[34]

Co

rnea

lC

oll

agen

Cro

ss-L

ink

ing

for

Ect

asia

Aft

erE

xci

mer

Las

er

Ref

ract

ive

Su

rger

y:

1-Y

ear

Res

ult

s

JR

efra

ctS

urg

22

:1–

12

13

eyes

Mea

n1

2m

on

ths

NS

C1

lin

eM

ean

1.4

NS

No

tst

atis

tica

lly

sig

nifi

can

t

Int Ophthalmol (2011) 31:309–319 313

123

Some researchers have attempted to combine CXL

with other therapeutic interventions for KC. Kymio-

nis et al. [41] demonstrated customized topography-

guided PRK followed by CXL with promising results,

while Kanellopoulos [42] in a study of 325 kerato-

conic eyes concluded that same-day simultaneous

topography-guided PRK and CXL appears to be

superior to sequential CXL with later PRK in the

visual rehabilitation of progressing KC. Also, Kymi-

onis et al. [43] performed conductive keratoplasty

(CK) followed by corneal collagen cross-linking

(CXL) in two patients with KC. Immediately after

CK, a significant corneal topographic improvement

was observed. The CK effect regressed 3 months

postoperatively and remained unchanged until the

sixth postoperative month in both patients. Kambu-

roglu et al. [44] reported the postoperative course and

refractive results of a 27-year-old patient in whom

intracorneal ring segment implantations (INTACS)

and CXL treatments were applied. Ertan et al. [45]

treated 25 eyes of 17 KC patients using INTACS

implantation with subsequent CXL treatment and

concluded that collagen cross-linking has an additive

effect on INTACS implantation in these eyes. This

combined treatment method may thus be considered

as an enhancement/stabilizing procedure.

Corneal collagen cross-linking and refractive

surgery

It is known that both laser-assisted in situ keratom-

ileusis (LASIK) and PRK modify corneal stability

and overall tissue strength. Researchers suggested

that regression after refractive surgery might be

reduced by means of CXL. This treatment appears to

be particularly applicable to patients who are sus-

pected of having forme fruste KC or those with a

correction that exceeds 8 D. These patients may

benefit from stiffening of the cornea because kera-

tectasia would not develop.

Early CXL treatments of iatrogenic keratectasia

after LASIK performed at the University of Dresden

seem to be effective in preventing further progression

of the central KC. Hafezi et al. [46] reported that

CXL arrested and/or partially reversed keratectasia in

10 patients over a postoperative follow-up of up to

25 months, as demonstrated by preoperative and

postoperative corneal topography and a reduction in

maximum keratometric readings. Similar results were

reported by Vinciguerra et al. [47] (Table 1).

With regard to the influence of CXL on excimer

laser surgery, CXL reduces the amount of refractive

change after myopic LASIK. While laser ablation

rate is unaffected, CXL results in an increased flap

thickness. This suggests a need for adjustment of the

microkeratome and laser parameters for LASIK after

CXL and indirectly endorses the theory of a direct

stiffening effect of CXL [48].

Corneal collagen cross-linking in pellucid

marginal degeneration

Kymionis et al. [49] performed simultaneous photo-

refractive keratectomy and CXL in a 34-year-old

woman with progressive pellucid marginal corneal

degeneration in both eyes. Twelve months postoper-

atively, BCVA improved from 20/50 to 20/25 and

20/63 to 20/32 in the right and left eye, respectively.

Corneal topography revealed significant improve-

ment in both eyes. Also, Spadea [50] reported the

results of CXL in a patient affected by pellucid

marginal degeneration. Corrected distance visual

acuity improved from 20/200 to 20/63 at 3 months

and was stable through the 12 month interval,

keratometric astigmatism showed a 1.4 D reduction

and the power of ectasia apex decreased from 82 to

78 D.

Corneal collagen cross-linking in corneal oedema

CXL has been shown to have an anti-oedematous

effect in the cornea. Wollensak et al. [51] therefore

examined whether this effect can be used for the

treatment of bullous keratopathy. This clinical inter-

ventional case series included three patients with

bullous keratopathy due to pseudophakia, corneal

transplant rejection and Fuchs’ endothelial dystrophy.

Corneal thickness was reduced. The bullous changes

of the epithelium were markedly improved, resulting

in loss of pain and discomfort. Visual acuity was

significantly improved in the case without prior

stromal scarring. Similar results were reported by

Krueger et al. [52], who performed staged intrastro-

mal delivery of riboflavin with UVA cross-linking in

a case of advanced bullous keratopathy, while

314 Int Ophthalmol (2011) 31:309–319

123

Ghanem et al. [53] in a study of 14 pseudophakic

bullous keratopathy eyes observed that corneal CXL

significantly improved corneal transparency, corneal

thickness and ocular pain 1 month postoperatively;

however, it did not seem to have a long-lasting effect

in decreasing pain and maintaining corneal transpar-

ency. The Ehlers et al. [54] study also showed a

potential application of collagen cross-linking in the

management of patients with corneal oedema, while

Cordeiro Barbosa et al. [55] investigated the pachy-

metric changes of collagen cross-linking in corneas

with oedema because of endothelial dysfunction and

concluded that CXL could be a safe procedure and

potential therapeutic alternative for the treatment of

corneal oedema. Recently, Hafezi et al. [56] reported

that modified CXL reduces corneal oedema and

diurnal visual fluctuations in Fuchs’ dystrophy, while

Bottos et al. [57] reported in their study that cross-

linked corneas had a pronounced anterior zone of

organized collagen fibres and even the treated

corneas with advanced bullous keratopathy and

stromal fibrosis had histological evidence of collagen

fibre organization, but this effect seemed to be

decreased compared with corneas in initial stages of

the disease. Despite the encouraging results, longer

follow-up is necessary to confirm the long-term

impact of CXL in corneal oedema.

Corneal collagen cross-linking in infectious

keratitis

In 2008, Martins et al. [58] demonstrated the

antimicrobial properties of CXL against common

pathogens. A group of bacteria was tested: Pseudo-

monas aeruginosa (PA), Staphylococcus aureus

(SA), Staphylococcus epidermidis (SE), methicillin-

resistant S. aureus (MRSA), multidrug-resistant P.

aeruginosa (MDRPA), drug-resistant Streptococcus

pneumoniae (DRSP) and Candida albicans (CA).

Riboflavin/UVA was effective against SA, SE, PA,

MRSA, MDRPA, and DRSP, but was ineffective

against CA.

Iseli et al. [59] evaluated the efficacy of CXL for

treating infectious melting keratitis. Five patients

with infectious keratitis associated with corneal

melting were treated with CXL when the infection

did not respond to systemic and topical antibiotic

therapy. Follow-up after cross-linking ranged from 1

to 9 months. In all cases, the progression of corneal

melting was halted after CXL treatment. Emergency

keratoplasty was not necessary in any of the five

cases presented.

Moreover, Micelli-Ferrari et al. [60] described a

case of keratitis caused by the Gram-negative Esch-

erichia coli treated with CXL with outstanding

outcome. Similar results were also reported by Moren

et al. [61], who treated an infectious keratitis using

CXL. These cases show the positive effects of CXL

on presumed infectious keratitis with a satisfactory

final visual outcome. This may be a promising new

treatment for keratitis, although this remains to be

elucidated in detail in future studies. Until more data

are available this treatment should only be considered

in therapy-refractive keratitis or ulceration and not in

the first line of defence, since it may have cytotoxic

side effects.

Corneal collagen cross-linking in complicated

bullous keratopathy with ulcerative keratitis

CXL’s antimicrobial and anti-oedematous properties

were demonstrated by Kozobolis et al. [62] in their

report of two patients with combined bullous kera-

topathy and ulcerative keratitis, resistant to conven-

tional treatments. Both patients presented significant

improvement of their vision-threatening corneal

ulcer, corneal oedema and BCVA for a follow-up

period of 2 months. Similar results were reported by

Ehlers et al. [63].

Risks and side effects

Riboflavin is generally regarded as a safe compound

since it is a vitamin ingested in normal diets and an

omnipresent molecule in biological systems. It is also

assumed that any residual riboflavin and any photo-

products produced during the treatment do not

present any hazardous risks. On the other hand, UV

represents a potential danger to the human eye. It is

well known that UV-induced photochemical damage,

like sunburn or photokeratitis, is caused by UVB

light. In the cornea UVB light (290–320 nm) is

mainly absorbed by the corneal epithelium [64, 65].

CXL uses a small peak-like sector of the UVA

spectrum (370 nm). UVA absorption in the cornea is

Int Ophthalmol (2011) 31:309–319 315

123

increased massively during the cross-linking proce-

dure due to the photosensitizer riboflavin, resulting in a

UVA transmission of only 7% across the cornea [8].

Experiments in rabbits indicated that the cytotoxic

level for the corneal endothelium was 0.36 mW/cm2.

The standard CXL procedure cannot provide this

cytotoxic level unless the central corneal thickness

(CCT) is less than 400 lm [66, 67]. In general, UVA is

absorbed mainly by the crystalline lens, which also

contains endogenous riboflavin and other photosensi-

tizers leading to cross-linking of crystallines [2]. This

intrinsic system protects the retina and for this reason a

UV absorber is usually incorporated into intraocular

lenses. With the standard CXL procedure, the lens only

receives 0.65 J/cm2 which is far below the cataract-

ogenous level of 70 J/cm2 [65]. In the rhesus monkey,

retinal damage with complete loss of the photoreceptor

layer was reported at a threshold level of 81 mW/cm2;

however, such extreme values are incompatible with

the standard treatment protocol [68].

On the other hand, after CXL, the stroma is

depopulated of keratocytes. HRT II-RCM confocal

microscopy indicated that the reduction in anterior

and intermediate stromal keratocytes is followed by

gradual repopulation which will be concluded within

6 months. The UVA induces oxygen radicals which

in turn induce covalent cross-linking of all kinds of

proteins. The main and almost exclusive protein of

the cornea is collagen type I, so that in the cornea the

effect is focused on collagen. UVA can induce DNA

and RNA lesions, an effect which is used for the

disinfection of water etc. and also for the sterilization

of aphaeresis blood products, efficiently killing

viruses, bacteria and other pathogens. These cyto-

toxic DNA lesions are also the reason for the

cytotoxic effect of the treatment on corneal kerato-

cytes. As long as the cornea treated has a minimum

thickness of 400 lm (as recommended), the corneal

endothelium will not experience damage, nor will

deeper structures such as lens and retina. The light

source should provide a homogenous irradiance,

avoiding hot spots [69–73].

Nevertheless, Mazzotta et al. [74] presented two

cases, studied through in vivo confocal microscopy,

with stage III KC that developed stromal haze after

the cross-linking treatment, Kymionis et al. [75]

reported the development of posterior linear stromal

haze after simultaneous PRK followed by CXL and

Raiskup et al. [76], after an extensive one-year study

of 163 eyes, concluded that advanced KC should be

considered at higher risk of haze development after

CXL due to low corneal thickness and high corneal

curvature.

In addition, Koppen et al. [77] reported four cases

of keratitis and corneal scarring from a total of 117

eyes treated with CXL and Sharma et al. [78]

presented a case report of Pseudomonas keratitis

after collagen cross-linking for KC.

Apart from the above-mentioned complications,

Gokhale et al. [79] presented a case of acute corneal

melt with perforation in a patient with KC after

collagen cross-linking treatment and the use of

topical diclofenac and proparacaine eyedrops. The

use of diclofenac sodium and proparacaine eyedrops

after surgery was possibly responsible for the corneal

melt in this patient, so patients who have undergone

cross-linking treatment should be observed closely

until the corneal epithelium heals completely.

In addition, a case of advancing KC treated with

CXL complicated with sterile infiltrates was pre-

sented by Mangioris et al. [80]. This complication

may be an individual hypersensitivity reaction to the

riboflavin or UVA light in the anterior stroma.

At this point we should mention that Goldich et al.

[81] eported that CXL does not cause damage to the

corneal endothelium and central retina and Koller

et al. [82] made a significant effort to evaluate the

complication rate of CXL for primary keratectasia

and to develop recommendations for avoiding com-

plications. Their results indicate that changing the

inclusion criteria may significantly reduce the com-

plications and failures of CXL. A preoperative

maximum K reading of less than 58 D may reduce

the failure rate to less than 3%, and restricting patient

age to younger than 35 years may reduce the

complication rate to 1%.

Conclusion

CXL is a promising therapeutic intervention for

corneal tissue stabilization in diseases that manifest

with progressive keratectasia, like keratoconus. More-

over, CXL’s anti-oedematous and antimicrobial prop-

erties have been demonstrated in a series of studies,

suggesting its therapeutic indications in bullous

keratopathy and in infectious keratitis, as an adjuvant

treatment to conventional therapeutic modalities.

316 Int Ophthalmol (2011) 31:309–319

123

References

1. Wollensak G (2006) Crosslinking treatment of progressive

keratoconus: new hope. Curr Opin Ophthalmol 17:356–360

2. Krishna CM, Uppuluri S, Riesz P et al (1991) A study of

the photodynamic efficiencies of some eye lens constitu-

ents. Photochem Photobiol 54:51–58

3. Bottos KM, Dreyfuss JL, Regatieri CV et al (2008)

Immunofluorescence confocal microscopy of porcine

corneas following collagen cross-linking treatment with

riboflavin and ultraviolet A. J Refract Surg 24(7):

715–719

4. Samaras K, O’brart DP, Doutch J et al (2009) Effect of

epithelial retention and removal on riboflavin absorption in

porcine corneas. J Refract Surg 25(9):771–775

5. Søndergaard AP, Hjortdal J, Breitenbach T, Ivarsen A

(2010) Corneal distribution of riboflavin prior to collagen

cross-linking. Curr Eye Res 35(2):116–121

6. Wollensak G, Aurich H, Wirbelauer C, Sel S (2010) Sig-

nificance of the riboflavin film in corneal collagen cross-

linking. J Cataract Refract Surg 36(1):114–120

7. Wollensak G, Spoerl E, Seiler T (2003) Riboflavin/ultra-

violet-A-induced collagen crosslinking for the treatment of

keratoconus. Am J Ophthalmol 135:620–627

8. Wollensak G, Sporl E, Seiler T (2003) Treatment of ker-

atoconus by collagen crosslinking. Ophthalmologe 100:

44–49 (In German)

9. Boxer Wachler BS (2005) Corneal collagen cross-linking

with riboflavin. Cataract Refract Surg Today 2005:73–74

10. Kanellopoulos AJ (2009) Collagen cross-linking in early

keratoconus with riboflavin in a femtosecond laser-created

pocket: initial clinical results. J Refract Surg 25(11):

1034–1037

11. Leccisotti A, Islam T (2010) Transepithelial corneal col-

lagen cross-linking in keratoconus. J Refract Surg 25:1–7

12. Bakke EF, Stojanovic A, Chen X, Drolsum L (2009)

Penetration of riboflavin and postoperative pain in corneal

collagen crosslinking: excimer laser superficial versus

mechanical full-thickness epithelial removal. J Cataract

Refract Surg 35(8):1363–1366

13. Wollensak G, Iomdina E (2009) Long-term biomechanical

properties of rabbit cornea after photodynamic collagen

crosslinking. Acta Ophthalmol 87(1):48–51

14. Kling S, Remon L, Perez-Escudero A, Merayo-Lloves J,

Marcos S (2010) Corneal biomechanical changes after

collagen cross-linking from porcine eye inflation experi-

ments. Invest Ophthalmol Vis Sci 51(8):3961–3968

15. Wollensak G, Spoerl E, Seiler T (2003) Stress–strain

measurements of human and porcine corneas after ribo-

flavin–ultraviolet-A-induced cross-linking. J Cat Refr Surg

29:1780–1785

16. Spoerl E, Wollensak G, Dittert DD, Seiler T (2004)

Thermomechanical behavior of collagen-cross-linked por-

cine cornea. Ophthalmologica 218:136–140

17. Wollensak G, Wilsch M, Spoerl E, Seiler T (2004) Col-

lagen fiber diameter in the rabbit cornea after collagen

crosslinking by riboflavin/UVA. Cornea 23:503–507

18. Faulborn J, Dunker S, Bowald S (1998) Diabetic vitreop-

athy: findings using the celloidin embedding technique.

Ophthalmologica 212:369–376

19. Wollensak G, Spoerl E, Wilsch M, Seiler T (2004) Kera-

tocyte apoptosis after corneal collagen cross-linking using

riboflavin/UVA treatment. Cornea 23:43–49

20. Wollensak G, Spoerl E, Reber F, Seiler T (2004) Kerato-

cyte cytotoxicity of riboflavin/UVA-treatment in vitro. Eye

18:718–722

21. Wollensak G, Iomdina E, Dittert DD, Herbst H (2007)

Wound healing in the rabbit cornea after corneal collagen

cross-linking with riboflavin and UVA. Cornea 26(5):

600–605

22. Wollensak G, Aurich H, Pham DT, Wirbelauer C (2007)

Hydration behavior of porcine cornea crosslinked with

riboflavin and ultraviolet A. J Cataract Refract Surg

33(3):516–521

23. Schilde T, Kohlhaas M, Spoerl E, Pillunat LE (2008)

Enzymatic evidence of the depth dependence of stiffening

on riboflavin/UVA treated corneas. Ophthalmologe

105(2):165–169 (In German)

24. Kohlhaas M, Spoerl E, Schilde T et al (2006) Biome-

chanical evidence of the distribution of cross-links in

corneas treated with riboflavin, ultraviolet A light. J Cata-

ract Refract Surg 32:279–283

25. Wollensak G, Vogten H (2004) Microfistola corneale in

caso di cheratocono acuto. Boll Ocul 83:235–238

26. Muller LJ, Pels E, Vrensen G (2002) The specific archi-

tecture of the anterior stroma accounts for maintenance of

corneal curvature. Br J Ophthalmol 85:437–443

27. Spoerl E, Wollensak G, Seiler T (2004) Increased resis-

tance of crosslinked cornea against enzymatic digestion.

Curr Eye Res 29:35–40

28. Kaufman HE (2004) Strengthening the cornea. Cornea

23:432

29. Labetoulle M (2003) An alternative to corneal transplan-

tation in keratoconus treatment? J Fr Ophtalmol 26:

1097–1098 (In French)

30. Sandner D, Sporl E, Kohlhaas M et al (2004) Collagen

crosslinking by combined riboflavin/ultraviolet-A (UVA)

treatment can stop the progression of keratoconus

(Abstract). Invest Ophthalmol Vis Sci 2004:45

31. Caporossi A, Baiocchi S, Mazzotta C et al (2006) Para-

surgical therapy for keratoconus by riboflavin-ultraviolet

type A rays induced cross-linking of corneal collagen;

preliminary refractive results in an italian study. J Cataract

Refract Surg 32:837–845

32. Witting-Silva C, Whiting M, Lamoureux E et al (2008) A

randomized controlled trial of corneal collagen cross-

linking in progressive keratoconus: preliminary results.

J Refract Surg 24(7):720–725

33. Raiskup-Wolf F, Hoyer A, Spoerl E et al (2008) Collagen

crosslinking with riboflavin and ultraviolet-A light in ker-

atoconus: long-term results. J Cataract Refract Surg 34(5):

796–801

34. Vinciguerra P, Albe E, Trazza S, Seiler T, Epstein D

(2009) Intraoperative and postoperative effects of corneal

collagen cross-linking on progressive keratoconus. Arch

Ophthalmol 127(10):1258–1265

35. Caporossi A, Mazzotta C, Baiocchi S, Caporossi T (2010)

Long-term results of riboflavin ultraviolet a corneal col-

lagen cross-linking for keratoconus in Italy: the Siena eye

cross study. Am J Ophthalmol 149(4):585–593

Int Ophthalmol (2011) 31:309–319 317

123

36. Hoyer A, Raiskup-Wolf F, Sporl E, Pillunat LE (2009)

Collagen cross-linking with riboflavin and UVA light in

keratoconus. Results from Dresden. Ophthalmologe 106(2):

133–140

37. Doors M, Tahzib NG, Eggink FA, Berendschot TT,

Webers CA, Nuijts RM (2009) Use of anterior segment

optical coherence tomography to study corneal changes

after collagen cross-linking. Am J Ophthalmol 148(6):

844–851

38. Grewal DS, Brar GS, Jain R et al (2009) Corneal collagen

cross-linking using riboflavin and ultraviolet-a light for

keratoconus: one-year analysis using Scheimpflug imag-

ing. J Cataract Refract Surg 35(3):425–432

39. Vinciquerra P, Albe E, Trazza S et al (2009) Refractive,

topographic, tomographic and aberrometric analysis of

keratoconic eyes undergoing corneal cross-linking. Oph-

thalmology 116(3):369–378

40. Coskunseven E, Jankov MR II, Hafezi F (2009) Contra-

lateral eye study of corneal collagen crosslinking with

riboflavin and UVA irradiation in patients with keratoco-

nus. J Refract Surg 25(4):371–376

41. Kymionis GD, Kontadakis GA, Kounis GA et al (2009)

Simultaneous topography-guided PRK followed by corneal

collagen cross-linking for keratoconus. J Refract Surg

25(9):807–811

42. Kanellopoulos AJ (2009) Comparison of sequential vs

same-day simultaneous collagen cross-linking and topog-

raphy-guided PRK for treatment of keratoconus. J Refract

Surg 25(9):812–818

43. Kymionis GD, Kontadakis GA, Naoumidi TL et al (2010)

Conductive keratoplasty followed by collagen cross-link-

ing with riboflavin-UV-A in patients with keratoconus.

Cornea 29(2):239–243

44. Kamburoglu G, Ertan A (2008) Intacs implantation with

sequential collagen cross-linking treatment in postopera-

tive LASIK ectasia. J Refract Surg 24(7):726–729

45. Ertan A, Karacal H, Kamburoglu G (2009) Refractive and

topographic results of transepithelial cross-linking treat-

ment in eyes with intacs. Cornea 28(7):719–723

46. Hafezi F, Kanellopoulos J, Wiltfang R, Seiler T (2007)

Corneal collagen crosslinking with riboflavin and ultravi-

olet A to treat induced keratectasia after LASIK. J Cataract

Refract Surg 33(12):2035–2040

47. Vinciguerra P, Camesasca FI, Albe E, Trazza S (2009)

Corneal collagen cross-linking for ectasia after excimer laser

refractive surgery: 1 year results. J Refract Surg 22:1–12

48. Kampik D, Ralla B, Keller S, Hirschberg M, Friedl P,

Geerling G (2010) Influence of corneal collagen crosslink-

ing with riboflavin and ultraviolet-A irradiation on excimer

laser surgery. Invest Ophthalmol Vis Sci 51(8):3929–3934

49. Kymionis GD, Karavitaki AE, Kounis GA et al (2009)

Management of pellucid marginal corneal degeneration

with simultaneous customized photorefractive keratectomy

and collagen crosslinking. J Cataract Refract Surg 35(7):

1298–1301

50. Spadea L (2010) Corneal collagen cross-linking with

riboflavin and UVA irradiation in pellucid marginal

degeneration. J Refract Surg 26(5):375–377

51. Wollensak G, Aurich H, Wirbelauer C, Pham DT (2009)

Potential use of riboflavin/UVA crosslinking in bullous

keratopathy. Ophthalmic Res 41(2):114–117

52. Krueger RR, Ramos-Esteban JC, Kanellopoulos AJ (2008)

Staged intrastromal delivery of riboflavin with UVA cross-

linking in advanced bullous keratopathy: laboratory inves-

tigation and first clinical case. J Refract Surg 24(7):730–736

53. Ghanem RC, Santhiago MR, Berti TB, Thomaz S, Netto

MV (2010) Collagen crosslinking with riboflavin and

ultraviolet-A in eyes with pseudophakic bullous keratopa-

thy. J Cataract Refract Surg 36(2):273–276

54. Ehlers N, Hjortdal J (2008) Riboflavin-ultraviolet light

induced cross-linking in endothelial decompensation. Acta

Ophthalmol 86(5):549–551

55. Cordeiro Barbosa MM, Barbosa JB Jr, Hirai FE, Hofling-

Lima AL (2010) Effect of cross-linking on corneal thick-

ness in patients with corneal edema. Cornea 29(6):613–617

56. Hafezi F, Dejica P, Majo F (2010) Modified corneal col-

lagen crosslinking reduces corneal oedema and diurnal

visual fluctuations in Fuchs dystrophy. Br J Ophthalmol

94(5):660–661

57. Bottos KM, Hofling-Lima AL, Barbosa MC, Barbosa JB

Jr, Dreyfuss JL, Schor P, Nader HB (2010) Effect of col-

lagen cross-linking in stromal fibril organization in edem-

atous human corneas. Cornea 29(7):789–793

58. Martins SA, Combs JC, Noguera G et al (2008) Antimi-

crobial efficacy of riboflavin/UVA combination (365 nm)

in vitro for bacterial and fungal isolates: a potential new

treatment for infectious keratitis. Invest Ophthalmol Vis

Sci 49:3402–3408

59. Iseli HP, Thiel AM, Hafezi F et al (2008) Ultraviolet

A/riboflavin corneal cross-linking for infectious keratitis

associated with corneal melts. Cornea 27(5):590–594

60. Micelli Ferrari T, Leozappa M, Lorusso M et al (2009)

Escherichia coli keratitis treated with ultraviolet A/ribo-

flavin corneal cross-linking: a case report. Eur J Ophthal-

mol 19(2):295–297

61. Moren H, Malmsjo M, Mortensen J, Ohrstrom A (2010)

Riboflavin and ultraviolet a collagen crosslinking of the

cornea for the treatment of keratitis. Cornea 29(1):102–104

62. Kozobolis V, Labiris G, Gkika M, Sideroudi H, Kaloghi-

anni E, Papadopoulou D, Toufexis G (2010) UVA collagen

cross linking treatment of bullous keratopathy combined

with corneal ulcer. Cornea 29(2):235–238

63. Ehlers N, Hjortdal J, Nielsen K, Søndergaard A (2009)

Riboflavin-UVA treatment in the management of edema

and nonhealing ulcers of the cornea. J Refract Surg

25(9):803–806

64. Podskochy A (2004) Protective role of corneal epithelium

against ultraviolet radiation damage. Acta Ophthalmol

Scand 82:714–717

65. Pitts DG, Cullen AP, Hacker PD (1977) Ocular effects ofultraviolet radiation from 295–365 nm. Invest Ophthalmol

Vis Sci 16:932–939

66. Wollensak G, Spoerl E, Wilsch M, Seiler T (2003) Endo-

thelial cell damage after riboflavin-ultraviolet-A treatment

in the rabbit. J Cataract Refract Surg 29:1786–1790

67. Wollensak G, Sporl E, Reber F et al (2003) Corneal

endothelial cytotoxicity of riboflavin/UVA treatment in

vitro. Ophthalmic Res 35:324–328

68. Zuclich JA (1989) Ultraviolet-induced photochemical

damage in ocular tissues. Health Phys 56:671–682

69. Mazzotta C, Balestrazzi A, Traversi C et al (2007) Treat-

ment of progressive keratoconus by riboflavin-UVA–

318 Int Ophthalmol (2011) 31:309–319

123

induced cross-linking of corneal collagen: ultrastructural

analysis by Heidelberg retinal tomograph II in vivo con-

focal microscopy in humans. Cornea 26:390–397

70. Spoerl E, Mrochen M, Sliney D et al (2007) Safety of

UVA-riboflavin cross-linking of the cornea. Cornea 26:

385–389

71. Avitabile T, Marano F, Uva MG, Reibaldi A (1997)

Evaluation of central and peripheral corneal thickness with

ultrasound biomicroscopy in normal and keratoconic eyes.

Cornea 16:639–644

72. Mencucci R, Marini M, Paladini I, Sarchielli E, Sgambati

E, Menchini U, Vannelli GB (2010) Effects of riboflavin/

UVA corneal cross-linking on keratocytes and collagen

fibres in human cornea. Clin Exp Ophthalmol 38(1):49–56

73. Esquenazi S, He J, Li N, Bazan HE (2010) Immunofluor-

esence of rabbit corneas after collagen cross-linking

treatment with riboflavin and ultraviolet A. Cornea

29(4):412–417

74. Mazzotta C, Balestrazzi A, Baiocchi S, Traversi C, Ca-

porossi A (2007) Stromal haze after combined riboflavin-

UVA corneal collagen cross-linking in keratoconus: in

vivo confocal microscopic evaluation. Clin Exp Ophthal-

mol 35(6):580–582

75. Kymionis GD, Portaliou DM, Diakonis VF, Kontadakis

GA, Krasia MS, Papadiamantis AG, Coskunseven E, Pal-

likaris A (2010) Posterior linear stromal haze formation

after simultaneous photorefractive keratectomy followed

by corneal collagen cross linking. Invest Ophthalmol Vis

Sci 51(10):5030–5033

76. Raiskup F, Hoyer A, Spoerl E (2009) Permanent corneal

haze after riboflavin-UVA-induced cross-linking in kera-

toconus. J Refract Surg 25(9):824–828

77. Koppen C, Vryghem JC, Gobin L, Tassignon MJ (2009)

Keratitis and corneal scarring after UVA/riboflavin cross-

linking for keratoconus. J Refract Surg 25(9):819–823

78. Sharma N, Maharana P, Singh G, Titiyal JS (2010) Pseu-

domonas keratitis after collagen crosslinking for kerato-

conus: case report and review of literature. J Cataract

Refract Surg 36(3):517–520

79. Gokhale NS, Vemuganti GK (2010) Diclofenac-induced

acute corneal melt after collagen crosslinking for kerato-

conus. Cornea 29(1):117–119

80. Mangioris GF, Papadopoulou DN, Balidis MO, Poulas JL,

Papadopoulos NT, Seiler T (2010) Corneal infiltrates after

corneal collagen cross-linking. J Refract Surg 28:1–3

81. Goldich Y, Marcovich AL, Barkana Y, Avni I, Zadok D

(2010) Safety of corneal collagen cross-linking with UV-A

and riboflavin in progressive keratoconus. Cornea

29(4):409–411

82. Koller T, Mrochen M, Seiler T (2009) Complication and

failure rates after corneal crosslinking. J Cataract Refract

Surg 35(8):1358–1362

Int Ophthalmol (2011) 31:309–319 319

123