PDFlib PLOP: PDF Linearization, Optimization, Protection

Page inserted by evaluation versionwww.pdflib.com – sales@pdflib.com

Clinical and Surgical Technique

Treatment of recurrent corneal erosion syndrome using thecombination of oral doxycycline and topical corticosteroidLouis Wang MBBS, Hughie Tsang MBBS and Minas Coroneo MD FRANZCODepartment of Ophthalmology, Prince of Wales Hospital, Randwick, New South Wales, Australia

ABSTRACT

Recurrent corneal erosion syndrome is a debilitating conditionfor which there are many treatment options with varying ratesof success. One treatment of interest in recent years is thecombination of oral doxycycline and topical corticosteroids,both of which have been shown to inhibit key metalloprotein-ases important to disease pathogenesis.To assess the efficacyof this treatment, we conducted a retrospective single-observer case series involving all patients with recurrentcorneal erosion syndrome who were treated at a community-based clinic with oral doxycycline and topical corticosteroidbetween January 2000 and July 2007. Twenty-one patientswere identified.All received oral doxycycline 50 mg twice dailyand topical fluoromethalone 0.1% three times daily for at least4 weeks.At 8 weeks post commencement of treatment, 15/21patients (71%) were symptom free. All but one of thesepatients reported an improvement in symptoms. Of thosepatients not lost to follow up, 15/18 patients (83%) and 11/15patients (73%) denied any symptoms suggestive of relapse at6 and 12 months, respectively.Among the patients in remissionwas one who had responded poorly to other treatmentsincluding ocular lubricants, epithelial debridement, serum eye-drops, anterior stromal puncture, and phototherapeutickeratectomy. Treatment of recurrent corneal erosion syn-drome with the combination of oral doxycycline and topicalcorticosteroid is effective. It may help patients with recurrentcorneal erosion syndrome who have failed other forms oftreatment.This non-invasive treatment modality should also beconsidered as the first treatment option when conservativemanagement with ocular lubricants fails.

Key words: cornea, corneal erosion, corticosteroid,doxycycline, metallo-proteinase.

INTRODUCTION

Recurrent corneal erosion syndrome (RCES) was first men-tioned in 1872 by Hansen, who aptly described it as an‘intermittent neuralgic vesicular keratitis’.1 Typical symptomsinclude recurrent episodes of early morning ocular surfacepain, photophobia and lacrimation. The disorder typicallyfollows a shallow corneal injury, but may be non-traumaticin origin. Common clinical signs include loosely elevatedepithelium, epithelial microcysts, corneal epithelial defects,stromal infiltrates and opacities. Epithelial basement mem-brane dystrophy may predispose some patients to RCES. Auseful sign of epithelial basement membrane dystrophy is thepresence of corneal valance (Shahinian’s sign), which can beseen as a ‘scalloped line of tear film thinning’ during slit-lampexamination under cobalt blue filter and fluorosceinstaining.2 In the majority of cases, RCES is successfullymanaged by conservative management such as patching andlubrication. However, a significant percentage of these casesare recalcitrant, and painful episodes recur despite treatmentwith a variety of alternatives such as bandage contact lens,colloidal hyperosmotic solution, anterior stromal punctureand phototherapeutic keratectomy.

In recent years, doxycycline had been found to showinhibitory actions on metalloproteinases. Matrix metallopro-teinases are known to be upregulated in the cornea inpatients with RCES.3–5 Based on these findings, some clini-cians have started using combined corticosteroid and doxy-cycline for the treatment of RCES.6 As many of the othertreatment alternatives for recalcitrant RCES are invasive withvarying degrees of efficacy, we conducted a retrospectivecase series analysing the efficacy of this treatment to helpclinicians decide on the applicability of this treatment whenconservative lubrication fails.

METHODS

A retrospective case series comprising of all patients withRCES who were treated with oral doxycycline and

� Correspondence: Professor Minas T Coroneo, 2 St. Pauls Street, Randwick, NSW 2031, Australia. E-mail: m.coroneo@unsw.edu.au

Received 4 December 2006; accepted 2 November 2007.

Clinical and Experimental Ophthalmology 2007; 36: 8–12doi: 10.1111/j.1442-9071.2007.01648.x

© 2008 The AuthorsJournal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists

corticosteroid (fluoromethalone) eye drops in a community-based ophthalmology clinic from January 2000 to July 2007was collated using a computer database search.

Inclusion criteria for this case series

1 Clinical picture characteristic of RCES, with eitherA History of recurrent episodic eye pain, particularly

during the night or upon waking, orB Documented evidence of at least one previous

episode of minor and major erosion (i.e. ulceration),together with:

2 A recent episode of characteristic eye pain, and eitherA Evidence of minor erosion or major erosion (i.e.

ulceration) on slit-lamp examination at clinic, orB Evidence of a healing erosion and documentation

of recent slit-lamp evidence of erosion on a referralletter from another referring medical practitioner oroptometrist

Twenty-one were found to have been commenced on oraldoxycycline 50 mg twice daily and topical corticosteroiddrops (fluoromethalone 0.1% three times daily) for at least4 weeks. Each patient was assessed systematically for signs ofblepharitis, dry eye and epithelial basement membrane dys-trophy prior to commencement of treatment. Patients werespecifically asked about symptoms of early morning ocularpain. The exact frequency of symptoms in the months priorto review was not recorded at each visit as this would havebeen subject to significant reporter error. Patients insteadwere asked at the 8-week review whether or not they felt thattheir symptom frequency or severity had improved.

Standard practice was to implement this treatment for4 weeks, followed by a repeat review at the clinic at 8 weeks.Treatment, however, was continued in those showing slowersigns of improvement or in those considered at high risk ofearly relapse. Patients were then advised to continue takinglong-term ocular lubricants at night and were given theoption to return after 1 year for review. Patients were alsoinstructed to return immediately to the clinic if they experi-enced a recurrence in eye pain. A patient was considered tohave had a relapse if they complained of any further episodesof recurrent eye pain on waking or if there was slit-lampevidence of major or minor erosions.

Patient demographics

Twenty-one patients were identified in our case series.Women comprised 62% of the study population (13 female,8 male). The median age at commencement of treatment was54 years (range 20–71 years). A pre-existing corneal epithe-lial basement membrane dystrophy was present in threepatients (Patients 3, 11 and 17). One patient had epider-molysis bullosa (Patient 16). Eight patients (38%) had ahistory of previous ocular trauma. Of the remaining 13patients without a history of ocular trauma, five had evidenceof blepharitis, two had Sjogren’s syndrome, one had epider-

molysis bullosa, and two had epithelial basement membranedystrophy. One patient (Patient 8) suffered from a particu-larly recalcitrant form of recurrent corneal erosion. Prior tohis referral, he had been treated with ocular lubricants, fol-lowed later by epithelial debridement, phototherapeutickeratectomy on two separate occasions, as well as anteriorstromal puncture. These treatments all had very little effecton his symptoms and disease.

RESULTS

Our results have been tabulated (see Table 1). At the 8-weekfollow-up appointment, 15 of the 21 patients (71%) weresymptom free. The remaining six patients (29%) had occa-sional pain on waking in the morning. All but one of thesepatients (Patient 1) reported a subjective improvement insymptoms and a decrease in frequency and severity of oculardiscomfort. All patients also demonstrated a remarkableimprovement in ocular findings at slit-lamp examination at8 weeks. Seventeen of the 21 patients (81%) had no cornealstaining on slit-lamp examination and the remaining fourpatients (19%) were found to have minimal punctate stainingthat was much improved from the initial appearance prior tocommencing treatment.

Of the 21 patients included in this case series, 12 patientshad a follow-up duration greater than 1 year. Six did notreturn for further follow up after their clinic review at8 weeks. Of these six patients, three were subsequently ableto be contacted by telephone to assess whether they had anysymptoms suggestive of relapse (e.g. recurrence in eye pain,particularly at night or upon waking) or improvement insymptoms in the time following the cessation of treatment.They all denied experiencing any of the above symptoms, orthat they had been subsequently diagnosed with a recurrenceof corneal erosion by another medical practitioner oroptometrist. The results of these three patients are indicatedin Table 1 by the suffix (T). In a further three patients, lessthan 1 year had passed since commencement of treatment.These patients were also contacted by telephone 6 monthspost commencement of treatment and were assessed as towhether they also had symptoms suggestive of relapse.These patients also denied any symptoms suggestive ofrecurrence.

At 6 months post commencement of treatment, 15 of the18 patients (83%) were relapse free (three were lost to followup). At 1 year post commencement of treatment, there wasno evidence of relapse in 11 of the 15 patients (73%) whohad a follow-up duration of at least 1 year, and all but onepatient had reported a subjective decrease in the frequencyand severity of ocular symptoms.

DISCUSSION

Conservative treatment with ocular lubricants is found to beeffective in approximately 60% of patients with RCES over aperiod of 4 years.7 Nevertheless, recalcitrant cases of RCESare not uncommon. Many treatment options, both non-

Treating recurrent corneal erosion syndrome 9

© 2008 The AuthorsJournal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists

Tabl

e1.

Base

line

char

acte

rist

ics

and

resu

lts

of21

pati

ents

trea

ted

wit

hor

aldo

xycy

clin

ean

dto

pica

lfluo

rom

etha

lone

for

recu

rren

tco

rnea

lero

sion

synd

rom

e

Pati

ent

Num

ber

Age

(yea

rs)

Gen

der

Eye

affe

cted

Ocu

lar

back

grou

ndM

echa

nism

oftr

aum

aPr

evio

usTr

eatm

ent

Trea

tmen

tdu

rati

on(w

eeks

)

Sym

ptom

sat

8w

eeks

Subj

ecti

veim

prov

emen

tsy

mpt

oms

(bot

hfr

eque

ncy

and

seve

rity

)du

ring

the

cour

seof

the

8w

eeks

Ocu

lar

findi

ngs

at8

wee

ks

Rel

apse

ofsy

mpt

oms

wit

hin

6m

onth

spo

stco

mm

ence

men

tof

trea

tmen

t

Rel

apse

ofsy

mpt

oms

wit

hin

1ye

arpo

stco

mm

ence

men

tof

trea

tmen

t

Subj

ecti

veim

prov

emen

tin

sym

ptom

sbo

thfr

eque

ncy

and

seve

rity

at1

year

post

trea

tmen

tco

mpa

red

wit

hpr

ior

totr

eatm

ent

Con

tinu

edus

eof

noct

eoc

ular

lubr

ican

tsat

1ye

arpo

sttr

eatm

ent

156

FL

Behc

et’s

dise

ase

Sjog

ren’

ssy

ndro

me

Unk

now

nD

ebri

dem

ent,

ocul

arlu

bric

ants

ASP

18O

ccas

iona

lpai

non

wak

ing

No

chan

geN

AD

Yes

at4

wee

ksYe

sat

4w

eeks

No

chan

geYe

s

257

FL

Ros

acea

,ble

phar

itis

Unk

now

nO

cula

rlu

bric

ants

seru

m-b

ased

eye

drop

s

4N

one

Yes

Min

imal

evid

ence

ofpu

ncta

test

aini

ng/

reso

luti

onof

intr

aepi

thel

ial

vesi

cles

No

No

Yes

Yes

357

ML

Map

dotc

orne

aldy

stro

phy

Gla

sses

inju

ryO

cula

rlu

bric

ants

4N

one

Yes

Min

imal

evid

ence

ofpu

ncta

test

aini

ng

No

Yes

at50

wee

ksYe

sYe

s

455

FL

Nil

Abr

asio

nw

ith

Xm

astr

eeD

ebri

dem

ent

ocul

arlu

bric

ants

4O

ccas

iona

lpai

non

wak

ing

Yes

NA

DYe

sat

20w

eeks

Yes

at20

wee

ksYe

sYe

s

557

ML

Bell’

spa

lsy

blep

hari

tis

Unk

now

nLi

dhy

gien

eO

cula

rlu

bric

ants

4N

one

Yes

NA

DN

oN

oYe

sYe

s

636

FL

Blep

hari

tis

Unk

now

nN

il4

Non

eYe

sM

inor

punc

tate

stai

ning

infe

ri-

orly

No

(T)

No

(T)

Yes

Yes

757

ML

+R

Blep

hari

tis

Unk

now

nO

cula

rlu

bric

ants

4N

one

Yes

NA

DN

oN

oYe

sYe

s8

71M

L+

RBl

epha

riti

s,flo

ppy

eyel

idsy

ndro

me

Stic

kD

ebri

dem

ent,

ocul

arlu

bric

ants

PTK

(¥2)

,ASP

78O

ccas

iona

lpai

non

wak

ing

Yes

Red

uced

punc

tate

stai

ning

No

No

Yes

Yes

958

FL

Nil

Unk

now

nN

il4

Non

eYe

sN

AD

No

(T)

No

(T)

Yes

No

1046

FL

+R

Nil

Unk

now

nPr

edso

l,oc

ular

lubr

ican

ts32

Occ

asio

nalp

ain

onw

akin

gYe

sN

AD

No

No

Yes

Yes

1171

FL

+R

Cor

neal

dyst

roph

y?t

ype

unsp

ecifi

ed

Unk

now

nO

cula

rlu

bric

ants

24O

ccas

iona

lpai

non

wak

ing

Yes

NA

DN

oN

oYe

sYe

s

1254

FL

+R

Sjog

ren’

ssy

ndro

me

Unk

now

nO

cula

rlu

bric

ants

,se

rum

-bas

edey

edr

ops

12O

ccas

iona

lpai

non

wak

ing

No

NA

DN

oN

oN

oYe

s

1352

MR

Nil

Bran

chin

jury

wit

hve

geta

ble

mat

ter

Nil

8N

one

Yes

NA

DN

o(T

)N

o(T

)Ye

sYe

s

1464

FR

Gla

ucom

aFi

nger

iney

eO

cula

rlu

bric

ants

and

anti

biot

icey

edr

ops

20O

ccas

iona

lpai

non

wak

ing

Yes

NA

DN

oN

oYe

sYe

s

1524

ML

Nil

Unk

now

nSo

ftba

ndag

eco

ntac

tlen

s,oc

ular

lubr

ican

ts

8N

one

Yes

NA

DYe

sat

12w

eeks

Yes

at12

wee

ksYe

sYe

s

1620

FL

Epid

erm

olys

isbu

llosa

Unk

now

nO

cula

rlu

bric

ants

20†

Non

eYe

sN

AD

No

F/U

<1

year

F/U

<1

year

F/U

<1

year

1740

MR

Map

dotc

orne

aldy

stro

phy

Unk

now

nA

cycl

ovir

for

susp

ecte

dan

ddi

agno

sed

herp

etic

eye

dise

ase

4N

one

Yes

NA

DN

oF/

U<

1ye

arF/

U<

1ye

arF/

U<

1ye

ar

1824

FL

Blep

hari

tis

Unk

now

nO

cula

rlu

bric

ants

4N

one

Yes

NA

DN

oF/

U<

1ye

arF/

U<

1ye

arF/

U<

1ye

ar19

71M

RN

ilTr

eebr

anch

Ocu

lar

lubr

ican

ts4

Non

eYe

sN

AD

Lost

toF/

ULo

stto

F/U

Lost

toF/

ULo

stto

F/U

2052

FL

Nil

Pape

rcu

tN

il4

Non

eYe

sN

AD

Lost

toF/

ULo

stto

F/U

Lost

toF/

ULo

stto

F/U

2151

FR

Nil

Potp

lant

Nil

4N

one

Yes

NA

DLo

stto

F/U

Lost

toF/

ULo

stto

F/U

Lost

toF/

U

† Pati

ent

rece

ivin

gon

goin

gtr

eatm

ent.

ASP

,ant

erio

rst

rom

alpu

nctu

re;F

/U,F

ollo

wup

;NA

D,n

oab

norm

alit

yde

tect

edon

corn

eals

tain

ing;

PTK

,pho

toth

erap

euti

cke

rate

ctom

y;(T

),pa

tien

tw

asfo

llow

edup

byte

leph

one.

10 Wang et al.

© 2008 The AuthorsJournal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists

surgical (e.g. treatments directed against underlying Mei-bomian gland dysfunction, soft bandage contact lens,autologous serum and matrix metalloproteinase inhibitors)and surgical (epithelial debridement, anterior stromal punc-ture, phototherapeutic keratectomy and diamond burr super-ficial keratectomy), are available. Success rates, however,have varied considerably among the various publishedstudies.8 The recurrence rate after anterior cornea stromalpuncture has been reported to be up to 40%.9 The samestudy reported that epithelial basement membrane debride-ment had a recurrence rate of 18%.9 The recurrence rate ofphototherapeutic keratectomy is also highly variable, withrecurrence rates as high as 26–36%.10,11 As yet, there is nopublished study to date comparing long-term success rates ofcurrent non-surgical and surgical treatment modalities.

One rationale for treating RCES with doxycycline andcorticosteroid is for their inhibition of metalloproteinases.It is known that patients with RCES have raisedmetalloproteinase-2 activities in their corneal epithelium.3

Metalloproteinase-9 was also found in edges of non-healingcorneal ulcers.4 Metalloproteinases are involved in the deg-radation of corneal epithelial basement membrane priorto re-epithelization.5 Excessive activity of these enzymes,which cleave components of corneal epithelial adhesioncomplexes including collagen types IV, VII, lamininand fibronectin, can be harmful for the process ofre-epithelization. Doxycycline has the effect of inhibitingsome of these enzymatic actions. In human corneal epithe-lium culture, treatment with doxycycline had been found tomarkedly lessen pro-metalloproteinase-9 activities.6

The effect of doxycycline in RCES may also be a result ofan additional effect on meibomian gland dysfunction. Mei-bomian gland dysfunction has been thought to contribute tothe formation of recurrent corneal erosions by creating anenvironment where lipases from colonizing bacteria reactwith meibomian gland secretions to form an abnormal tearcomposition rich in toxic free fatty acids that can affectepithelial membrane integrity and healing.8 Doxycycline hasbeen demonstrated in a randomized control trial to have abeneficial effect in patients with chronic meibomian glanddysfunction.12 The effect of tetracyclines such as doxycylinehas been shown to be due to the inhibition of lipase produc-tion in Staphylococcus epidermidis rather than a reduction in thenumber of bacteria.13 Although the presence or absence ofmeibomian gland dysfunction was not routinely documentedin the medical notes of patients in our case series, it haspreviously been documented in a significant proportion ofpatients with RCES.14

Dursun et al. found that oral doxycycline and topical cor-ticosteroid was effective clinically in preventing relapse ofRCES in all seven of their patients.6 A randomized controltrial performed by Hope-Ross et al. found that over a24-week period, there was a significant decrease in thenumber of recurrences in patients using a combination of oraltetracycline (an antibiotic of the same antibacterial class asdoxycycline) and topical prednisolone, or tetracycline alone,compared with placebo.15 From our experience, the combi-

nation of oral doxycycline 50 mg twice a day with topicalfluoromethalone 0.1% resulted in the reduction of symptomsin all but one of our patients and afforded a 1-year relapse-free rate of 73%. Given that patients were less likely topresent for review (i.e. more likely to become lost to followup) if they were asymptomatic and relapse free, we believethat our true relapse-free rate may have been in fact higher.

Doxycycline is generally well tolerated and none of ourpatients reported any side-effects. However, common side-effects include nausea, diarrhoea, photosensitivity and otherskin reactions. Some rare side-effects include benign intrac-ranial hypertension, anaphylaxis and oesophageal ulceration.Doxycycline is also contraindicated in pregnant women inthe second and third trimesters, breastfeeding women and inyoung children due to the risk of discoloration of developingteeth, enamel dysplasia and effects on infant bone growth.16

Topical corticosteroid eyedrops also have side-effects includ-ing raised intraocular pressure and delayed corneal healing.This is usually reversible and the presence of raised intraocu-lar pressure was not noted in any of the patients in our caseseries at 8 weeks.

Despite these potential side-effects, treatment with doxy-cycline and topical corticosteroid may avoid the risks ofscarring or astigmatism as seen in anterior stromal puncture,or induced refractive error as seen in phototherapeutickeratectomy. We believe this makes the combination of oraldoxycycline and topical steroid treatment an attractive alter-native to surgical options and propose that its use should beconsidered in all patients where conservative measures suchas ocular lubricants have had a hitherto unsatisfactory effect.

Overall, we believe that the treatment of RCES withcombination doxycycline and topical steroid is promising.We recognize that further prospective comparison studies,with long-term follow up, are needed to confirm the efficacyof this treatment. However, our experience has been that thisreadily available treatment option offered at least a 70%short-term relapse-free rate, as well as a subjective improve-ment in symptoms and decrease in number of recurrences inthe majority of our patients.

REFERENCES

1. Hansen E. Om den intermittirende keratitis vesiculosa neural-gica af traumatisic oprindelse. Hosp Tidende 1872; 15: 201–3.

2. Shahinian L Jr. Corneal valance: a tear film pattern in map-dot-fingerprint corneal dystrophy. Ann Ophthalmol 1984; 16: 567,570–1.

3. Garrana RM, Zieske JD, Assouline M, Gipson IK. Matrix met-alloproteinases in epithelia from human recurrent cornealerosion. Invest Ophthalmol Vis Sci 1999; 40: 1266–70.

4. Fini ME, Parks WC, Rinehart WB et al. Role of matrix metallo-proteinases in failure to re-epithelialize after corneal injury. AmJ Pathol 1996; 149: 1287–302.

5. Di Girolamo N, Chui J, Coroneo M, Wakefield D. Pathogen-esis of pterygia: role of cytokines, growth factors, and matrixmetalloproteinases. Prog Retin Eye Res 2004; 23: 195–228.

Treating recurrent corneal erosion syndrome 11

© 2008 The AuthorsJournal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists

6. Dursun D, Kim MC, Solomon A, Pflugfelder SC. Treatment ofrecalcitrant recurrent corneal erosions with inhibitors of matrixmetalloproteinase-9, doxycycline and corticosteroids. Am JOphthalmol 2001; 132: 8–13.

7. Heyworth P, Morlet N, Rayner S, Hykin P, Dart J. Naturalhistory of recurrent erosion syndrome – a 4 year review of 117patients. Br J Ophthalmol 1998; 82: 26–8.

8. Ramamurthi S, Rahman MQ, Dutton GN, Ramaesh K. Patho-genesis, clinical features and management of recurrent cornealerosions. Eye 2006; 20: 635–44.

9. Reidy JJ, Paulus MP, Gona S. Recurrent erosions of the cornea:epidemiology and treatment. Cornea 2000; 19: 767–71.

10. Ohman L, Fagerholm P, Tengroth B. Treatment of recurrentcorneal erosions with the excimer laser. Acta Ophthalmol (Copenh)1994; 72: 461–3.

11. Baryla J, Pan YI, Hodge WG. Long term efficacy of photothera-peutic keratectomy on recurrent corneal erosion syndrome.Cornea 2006; 25: 1150–2.

12. Yoo SE, Lee DC, Chang MH. The effect of low dose doxycy-cline therapy in chronic meibomian gland dysfunction. Korean JOphthalmol 2005; 19: 258–63.

13. Dougherty JM, McCulley JP, Silvany RE, Meye DR. The roleof tetracycline in chronic blepharitis. Inhibition of lipase pro-duction in staphylococci. Invest Ophthalmol Vis Sci 1991; 32:2970–5.

14. Hope-Ross MW, Chell PB, Kervick GN, McDonnell PJ. Recur-rent corneal erosion: clinical features. Eye 1994; 8: 373–7.

15. Hope-Ross MW, Chell PB, Kervick GN, McDonnell PJ, JonesHS. Oral tetracycline in the treatment of recurrent cornealerosions. Eye 1994; 8: 384–8.

16. Doxycycline [monograph online]. In: Australian Medicines Handbook[Ebook]. Adelaide: Australian Medicines Handbook Pty Ltd,2007. Clinical Information Access Program. Accessed August2007. Available from: http://www.ciap.health.nsw.gov.au/

12 Wang et al.

© 2008 The AuthorsJournal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists