Introduction

Corticosteroid-induced ocular hyper-tension is a well-established clinicalobservation. In adults, topical admin-istration of 0.1% dexamethasone(DEX) for 4–6 weeks led to a rise ofintraocular pressure (IOP) > 6 mmHgin 35% of normal subjects (Armaly1963; Becker & Ballin 1965). In chil-dren, up to 80% can develop IOPrise of > 6 mmHg with 4- week topi-cal applications of 0.1% DEX twiceto four times daily, as shown in ourprevious study of Chinese childrenundergone strabismus operation (Lamet al. 2005). Corticosteroids chemi-cally engineered to reduce IOP elevat-ing potentials have reduced risk ofcausing IOP rise in adults and chil-dren, but the latter still display a rel-atively heightened ocular hypertensiveresponse.

Fluorometholone (FML; 21-deoxy-9-fluoro-6-methyl prednisolone), a pre-dnisolone derivative with structuralsimilarity to progesterone, caused IOPrise of > 5 mmHg in 8% of adulteyes (Akingbehin 1983). When appliedto children, FML caused a rise ofIOP ‡ 5 mmHg in 29% ofeyes (Fan et al. 2001). Rimexolone(RIM, 116-hydroxy-16a,17a-dimethyl-17-propionylandrostal,4-diene-3-one),

A rabbit model of age-dependantocular hypertensive response totopical corticosteroids

Yi Qin,1,*,� Shi Lam,1,* Gary Hin Fai Yam,1,* Kwong WaiChoy,1,2 David Tai Li Liu,1 Thomas Yee Hang Chiu,1,� Wai YingLi,1 Dennis Shun Chiu Lam,1 Chi Pui Pang1

and Dorothy Shu Ping Fan1

1Department of Ophthalmology and Visual Sciences, The Chinese University ofHong Kong, Hong Kong, China2Department of Obstetrics and Gynaecology, The Chinese University of HongKong, Hong Kong, China

ABSTRACT.

Objective: To investigate the ocular hypertensive response to topical dexa-

methasone (DEX), rimexolone (RIM), loteprednol etabonate (LOT) and fluo-

rometholone (FML) in rabbits of different ages.

Methods: Seventy-five rabbits of three age groups (7 weeks, 6 months and

1-year old) received topical administration of 0.1% DEX, 1% RIM, 0.5%

LOT, 0.1% FML or balanced salt solution four times daily for 1 month.

Intraocular pressure (IOP) was monitored at regular time intervals. After a

month, eyes were harvested for histological study with haematoxylin and eosin

(H&E), periodic acid Schiff and Masson trichrome staining. Trabecular mesh-

work changes were graded by masked ocular pathologists.

Results: Topical DEX caused the greatest increase in IOP, followed by RIM

and FML. LOT caused the least IOP increase. Similar pattern of IOP

response to the four corticosteroids was observed in the three studied age

groups. Young rabbits (7 week) were the most responsive to corticosteroids

among the age groups. Extracellular matrix thickening in the trabecular mesh-

work region and loss of trabecular meshwork cells were observed after DEX,

FML or RIM treatments.

Conclusion: Young rabbits are more susceptible to steroid induced increase in

IOP, even for milder steroids such as fluorometholone and rimexolone.

Key words: age dependent – glaucoma – ocular hypertension – topical corticosteroids

Acta Ophthalmol.ª 2010 The Authors

Acta Ophthalmologica ª 2010 Acta Ophthalmologica Scandinavica Foundation

doi: 10.1111/j.1755-3768.2010.02016.x

*These authors contributed equally.

Present addresses: �Beijing Tongren Eye Centre, Beijing Tongren Hospital, Capital Medical

University, Beijing Ophthalmol and Visual Science Key Lab, Beijing, China.�Hong Kong Baptist Hospital, Hong Kong, China.

Acta Ophthalmologica 2010

1

a nonfluorinated corticosteroid, ca-used IOP rise ‡ 5 mmHg in 1.5% ofadult eyes (Assil et al. 1997) and in50% of children eyes (Fan et al.2003). Loteprednol (LOT) is a ‘softsteroid’ that can undergo rapid meta-bolic inactivation after exerting itsaction. It was reported to cause IOPrise (> 10 mmHg) in 7% of adulteyes treated for various conditions(Dell et al. 1997; Friedlaender & Ho-wes 1997). The ocular hypertensiveeffect of LOT has not been evaluatedin children. A direct comparisonamong adult and paediatric subjectscontrolling the baseline ocular condi-tion is difficult to carry out in humansubjects. The age-dependent ocularhypertensive response to topical DEXhas been observed in rabbits (Knepperet al. 1978). In this study, we adopteda rabbit model of steroid-inducedocular hypertension to evaluate theage-dependant ocular hypertensiveresponse to different types of cortico-steroid and to investigate the possiblepathogenic mechanism of steroid-induced ocular hypertension.

Materials and Methods

Animal experiment

The experimental procedure wasadherent to the ARVO Statementfor Use of Animals in Ophthalmicand Vision Research and Guide forthe Care and Use of LaboratoryAnimals adopted and promulgatedby The University Animal EthicsCommittee. It was approved by Ani-mal Care Committee of The ChineseUniversity of Hong Kong, HongKong. A total of 75 male New Zea-land albino rabbits with age of7 weeks, 6 months and 1 year wereused. In each age group, rabbitswere randomized to receive topicaltreatment with different marketedcorticosteroids, namely 0.1% DEX(Maxidex; Alcon, Fort Worth, TX,USA) (n = 5), 1% RIM (Vexol;Alcon) (n = 5), 0.5% Loteprednoletabonate (LOT) (Lotemax; Baush &Lomb, Tampa, FL, USA) (n = 5)and 0.1% Fluorometholone (FML;Alcon) (n = 5) or balanced salt solu-tion (BSS; Alcon) (n = 5). The drugsuspensions were shaken for 10 sec-onds before use. One drop wasapplied to the right eye four timesdaily for a total of 30 days.

Monitoring of intraocular pressure

Intraocular pressure of both eyes wasrecorded at 11 am on day 0, 1, 2, 3and afterwards every 3 days duringtreatment period with Tono-Pen XL(Solan Ophthalmic Products, Jackson-ville, FL, USA) by a masked optome-trist. The mean of five measurementswas recorded. Peak IOP and time toreach peak IOP were determined. Sta-tistical analyses were conducted withStudent’s t-test and one-way analysisof variance (anova) test, and p value< 0.05 was considered statistically sig-nificant. The association of peak IOP,maximal IOP change with age andtype of corticosteroid was examined.

Histology

After topical treatment, rabbits werekilled by intravenous overdosing ofsodium pentobarbital (Alfasan, Woer-den, the Netherland), and their eyeswere harvested and bisected horizon-tally along the nasal-temporal line.The samples were fixed in 10% neutralbuffered formalin and processed forparaffin embedding. Ten consecutivesections (5-lm thick) were stained withhaematoxylin–eosin (H&E), Massontrichrome (MT) and periodic acid-Schiff (PAS) reagents. The sectionswere examined under light microscopy(DMRB; Leica, Wetzlar, Germany) by

two qualified ocular pathologistsmasked to the treatments. Trabecularmeshwork (TM) cellularity was gradedby nuclei counting as: 0 – < 5% lessthan control; 1 – 5–20% less than con-trol and 2 – > 20% less than control.Extracellular matrix thickening wasgraded by impressive judgment fromPAS and MT staining as 0 – no differ-ence from control; 1 – mild thickeningand 2 – marked thickening.

Result

Age effect on ocular hypertensive response

to different topical corticosteroids

The baseline IOP (± SD) in Group 1(7 weeks old), 2 (6 months old) and 3(1 year old) was 10.9 ± 1.1, 14.8 ±1.9 and 11.7 ± 3.9 mmHg, respec-tively; the baseline IOP of Group 2 rab-bits was significantly higher than theother two groups (p < 0.01, Student’st-test), while baseline IOPs of Group 1and 3 were not statistically different(p = 0.15, Student’s t-test). Table 1showed IOP profile at weekly intervalduring treatment, and the change ofIOP from baseline was plotted againsttime interval in Fig. 1. An age-depen-dant ocular hypertensive response wasobserved. In Group 1 rabbits, IOP wasconsistently 4–5 mmHg higher thanbaseline after 1 week of topical DEX,FML and RIM. In contrast, IOPs of

Table 1. Weekly IOP values (mean ± SD) of different treatment and age groups.

Week of IOP measurement p-value

(week 4

versus 0)0 1 2 3 4

BSS

Group 1 10.9 ± 0.5 11.9 ± 0.4 10.8 ± 0.3 11.9 ± 1.2 11.7 ± 0.4 0.07

Group 2 15.5 ± 0.7 16.6 ± 0.6 15.9 ± 0.5 14.8 ± 0.4 15.1 ± 0.6 0.33

Group 3 14.3 ± 2.5 15.0 ± 1.7 11.8 ± 1.0 14.1 ± 0.3 15.0 ± 1.2 0.55

DEX

Group 1 10.7 ± 1.4 14.4 ± 1.7 14.7 ± 1.5 14.5 ± 1.9 14.4 ± 1.2 < 0.01*

Group 2 13.4 ± 0.9 14.5 ± 0.7 14.8 ± 1.3 15.0 ± 1.2 15.0 ± 1.1 0.03*

Group 3 7.8 ± 1.2 7.7 ± 0.8 7.8 ± 1.8 8.9 ± 1.6 7.4 ± 0.8 0.57

FML

Group 1 10.8 ± 0.6 15.1 ± 1.9 14.9 ± 2.5 13.1 ± 0.7 13.3 ± 1.1 < 0.01*

Group 2 16.3 ± 2.2 16.7 ± 1.1 19.2 ± 0.4 14.7 ± 1.2 14.9 ± 1.4 0.29

Group 3 14.0 ± 0.7 13.8 ± 1.5 13.2 ± 1.6 15.0 ± 0.4 12.6 ± 0.9 0.03

RIM

Group 1 11.1 ± 1.6 14.8 ± 1.7 14.4 ± 2.0 14.1 ± 0.7 13.9 ± 1.1 0.05*

Group 2 15.5 ± 2.3 15.7 ± 0.9 14.6 ± 0.6 15.0 ± 0.6 14.8 ± 1.6 0.53

Group 3 7.1 ± 0.6 7.0 ± 1.4 8.8 ± 2.4 10.8 ± 1.2 8.3 ± 1.0 0.29

LOT

Group 1 12.8 ± 0.9 13.1 ± 0.7 14.3 ± 1.5 13.1 ± 0.7 13.4 ± 1.8 0.66

Group 2 15.8 ± 1.6 14.1 ± 0.8 13.3 ± 2.1 14.5 ± 0.8 15.0 ± 2.3 0.86

Group 3 15.5 ± 1.8 14.6 ± 1.1 13.7 ± 0.7 15.6 ± 2.1 13.0 ± 1.4 0.04

* means statistically significant.

Acta Ophthalmologica 2010

2

Group 2 and 3 rabbits were not signifi-cantly elevated after 4 weeks of topicalsteroid application. Table 2 showedthat Group 1 rabbits treated by DEXdisplayed the greatest IOP rise of5.6 mmHg, followed by FML with4.4 mmHg, RIM with 3.8 mmHg andLOT with 2.8 mmHg. When the samecorticosteroid was applied on older

rabbits, IOP rise was in general lessthan younger one.

Effect of topical corticosteroid on irido-

corneal angle structure in eyes of rabbits

Figure 2 showed that trabecular mesh-work cellularity of DEX, FML andRIM treated eyes was lower than

BSS-treated eyes with a correspond-ing increase in intercellular space(Fig. 3D–G, K, L). In contrast, BSScontrol eyes of different age group(n = 24) showed regular patterning ofcells with minimal intercellular space intrabecular meshwork region. With MTstaining, substantial thickening ofextracellular matrix materials (hyalini-zation) in trabecular meshwork regionwas observed in eyes after DEX, FMLor RIM treatments (Fig. 4D–F, H–J).Degree of such trabecular meshworkchanges was not age related. Notably,one RIM-treated 7-week-old rabbitexhibited marked trabecular mesh-work hyalinization; one DEX-treated6-month-old rabbit and one RIM-trea-ted 1-year-old rabbit showed markedtrabecular meshwork cell loss. In LOT-treated rabbits, trabecular meshworkchanges were significantly less com-pared to eyes treated by other steroids.

Discussion

We have demonstrated the age-depen-dent ocular hypertensive responses todifferent topical corticosteroids inalbino rabbit eyes. In our study, 0.1%DEX, FML and RIM induced IOPrise of 5.6, 4.4 and 3.8 mmHg, respec-tively, in 7-week-old rabbits. Althoughin a different species, our findingsagreed with that of (Knepper et al.1978), who reported that 0.1% DEXcaused 5 mmHg or more IOP rise in76% of young New Zealand Red rab-bits (8–10 weeks old) but not in olderrabbits (3 years old). Although theextent of melanin binding of differentcorticosteroids is unknown, a possibleeffect of significant binding could be areduced but prolonged drug activity.The use of albino rabbits provided usa more simplistic model for studyingcorticosteroid-induced IOP changes inyoung animals. We noted that theIOP of 6-month-old rabbits was sig-nificantly higher at baseline comparedwith the other two age groups. It washowever not significantly changedduring the course of steroid treatment.This is possibly because of chanceselection and may not carry any clini-cal significance.

The underlying pathophysiology ofcorticosteroid-induced ocular hyper-tension was suggested to be theimpairment of aqueous humour out-flow. Corticosteroid treatment was

(A)

(B)

Fig. 1. Intraocular pressure (IOP) changes in different treatment and age groups. (A) Compari-

son of IOP changes between different age groups under same treatment. IOP was elevated at

1- week treatment and remained elevated in 7-week-old rabbits treated with dexamethasone

(DEX), fluorometholone (FML) and rimexolone (RIM). No definite trend of IOP elevation in

6-month and 1-year-old rabbits was observed. (B) Comparison of IOP changes induced by dif-

ferent corticosteroids treatments under same age group. A differential IOP elevating effect

between Loteprednol and the other 3 (DEX, FML and RIM) was the most obvious in 7-week-

old rabbits.

Table 2. Peak IOP rise and time to reach peak IOP after topical corticosteroid application in

rabbits of different ages.

Corticosteroid

Group 1 (7-week) Group 2 (6-month) Group 3 (1-year)

Peak

IOP rise

(mmHg)

Day to

peak

IOP

(day)

Peak

IOP rise

(mmHg)

Day to

peak

IOP

(day)

Peak

IOP rise

(mmHg)

Day to

peak

IOP

(day)

DEX 5.6 ± 1.9 4 1.4 ± 1.2 11 0.2 ± 2.2 4

FML 4.4 ± 1.8 10 2.8 ± 1.4 11 < 0 –

RIM 3.8 ± 2.8 24 2.5 ± 2.4 7 1.7 ± 1.6 22

LOT 2.2 ± 1.2 19 0.4 ± 2.4 11 < 0 –

BSS 1.1 ± 0.7 26 0.6 ± 1.3 5 0.3 ± 0.4 9

Acta Ophthalmologica 2010

3

associated with trabecular meshworkchanges such as increased trabecularmeshwork cell volume, impaired tra-becular meshwork cell phagocyticfunction, altered trabecular meshworkcell cytoskeleton and altered extracel-lular matrix composition (Knepperet al. 1985; Wordinger & Clark 1999).Our findings of cell loss and sclerosisin the trabecular meshwork of eyestreated with DEX, FML and RIMagreed with the previous observations.Because of the inevitable distortion ofiris position during sample processingfor light microscopy, we did not com-mit to conclude on any minute differ-ence between the age groups.Trabecular meshwork cellularity wasrelatively unaffected by artefact oftissue processing and was shown todecrease linearly with age (Alvaradoet al. 1981); therefore, we adopted tra-becular meshwork cellularity gradingfor comparison of trabecular mesh-work alteration among different ageand treatment groups. On the otherhand, other age-related factors maycontribute to the exaggerated IOP ele-vation in young animals. As in human,the immature iridocorneal angle inyoung rabbits may be more susceptibleto the disruption of aqueous outflowgiven the same degree of structuralchange. An increased steroid respon-siveness in young rabbits can also bebecause of a higher corneal permeabil-ity in young rabbits (Ke et al. 1999).

Despite a general increase in sensi-tivity to topical corticosteroids inyoung rabbits, no IOP rise or trabecu-lar meshwork change was observed inLOT-treated eyes. An extra safety maybe conferred by the unique ‘soft ste-roid’ design of LOT compared to othercorticosteroids. A 17b-chloromethylester group in LOT molecule leads toits metabolic inactivation in the cornea(Bodor et al. 2000), resulting in rela-tively low aqueous and systemic bio-availability and thus a lower IOPelevating potential. Although the ‘sur-face activity’ of LOT favours its appli-cation in ocular surface inflammatoryconditions, its efficacy of controllingintraocular inflammation was reportedto be inferior to prednisolone.

There are limitations to considerwhen extrapolating our finding in rab-bits to humans. The aqueous outflowmechanism of rabbits is not identical tohuman, with the rabbit aqueous plexusin place of the human Schlemm’s canal.

(A)

(B)

Fig. 2. Frequency and grading of TM changes. (A) ECM thickening: 0 – no difference from

control; 1 – mild thickening; 2 – marked thickening; (B) TM cellularity: 0 – 0% to < 5% less

than control; 1 – 5–20% less; 2 – > 20% less.

6-month 1-year7-week

Control

DEX

FML

RIM

LOT

(O)(N)(M)

(L)(K)(J)

(I)(H)(G)

(F)(E)(D)

(C)(B)(A)

Fig. 3. Trabecular meshwork region stained by H&E. Increased intercellular space in trabecular

meshwork was observed more frequently in eyes treated with dexamethasone (DEX), FML and

rimexolone (RIM) comparing to BSS control. Loteprednol treatment did not result in trabecular

meshwork cell loss. (A)–(C) = Control; (D)–(F) = Dexamethasone (DEX); (G)–(I) = Fluoro-

metholone (FML); (J)–(L) = Rimexolone (RIM); (M)–(O) = Loteprednol etabonate (LOT).

Acta Ophthalmologica 2010

4

The albino rabbit model also does notaccount for melanin binding of corti-costeroids in human pigmented eyes.Finally, the significance and effect ofsystemic absorption of corticosteroidsis different between rabbits andhuman, as it is known that topical cor-ticosteroids can induce significantweight loss in rabbits.

In conclusion, we have reported analbino rabbit model of age-dependentocular hypertensive response to topi-cal corticosteroid with good resem-blance to clinical observation. DEX,FML and RIM caused increased IOPassociated with trabecular meshworkcell loss and ECM thickening. LOTdid not caused significant IOP rise or

histopathological change in youngrabbit eyes. This model hence pro-vided a reasonable platform for fur-ther investigations into the effect ofage on steroid responsiveness.

ReferencesAkingbehin AO (1983): Comparative study of

the intraocular pressure effects of fluorome-

tholone 0.1% versus dexamethasone 0.1%. Br

J Ophthalmol 67: 661–663.

Alvarado J, Murphy C, Polansky J & Juster R

(1981): Age-related changes in trabecular

meshwork cellularity. Invest Ophthalmol Vis

Sci 21: 714–727.

Armaly MF (1963): Effect of corticosteroids on

intraocular pressure and fluid dynamics I. The

effect of dexamethasone in the normal eye.

Arch Ophthalmol 70: 482–491.

Assil KK, Massry G, Lehmann R, Fox K &

Stewart R (1997): Control of ocular inflam-

mation after cataract extraction with rimexo-

lone 1% ophthalmic suspension. J Cataract

Refract Surg 23: 750–757.

Becker B & Ballin N (1965): Glaucoma and cor-

ticosteroid provocative testing. Arch Ophthal-

mol 74: 621–624.

Bodor N, Drustrup J & Wu W (2000): Effect of

cyclodextrins on the solubility and stability of

a novel soft corticosteroid, loteprednol eta-

bonate. Pharmazie 55: 206–209.

Dell SJ, Shulman DG, Lowry GM & Howes J

(1997): A controlled evaluation of the efficacy

and safety of loteprednol etabonate in the

prophylactic treatment of seasonal allergic

conjunctivitis. Loteprednol Allergic Conjunc-

tivitis Study Group. Am J Ophthalmol 123:

791–797.

Fan DSP, Ng JS & Lam DSC (2001): A pro-

spective study on ocular hypertensive and

antiinflammatory response to different dos-

ages of fluorometholone in children. Ophthal-

mology 108: 1973–1977.

Fan DSP, Yu CB, Chiu TY, Wong CY, Ng JS,

Pang CP & Lam DSC (2003): Ocular-hyper-

tensive and anti-inflammatory response to

rimexolone therapy in children. Arch Oph-

thalmol 121: 1716–1721.

Friedlaender MH & Howes J (1997): A dou-

ble-masked, placebo-controlled evaluation of

the efficacy and safety of loteprednol

etabonate in the treatment of giant papillary

conjunctivitis. The Loteprednol Etabonate

Giant Papillary Conjunctivitis Study Group

I. Am J Ophthalmol 123: 455–464.

Ke T, Clark AF & Gracy RW (1999): Age-

related permeability changes in rabbit

corneas. J Ocul Pharmacol Ther 15: 513–523.

Knepper PA, Breen M, Weinstein HG & Blacik

JL (1978): Intraocular pressure and glycos-

aminoglycan distribution in the rabbit eye:

effect of age and dexamethasone. Exp Eye

Res 27: 567–575.

Knepper PA, Collins JA & Frederick R (1985):

Effects of dexamethasone, progesterone, and

testosterone on IOP and GAGs in the rabbit

eye. Invest Ophthalmol Vis Sci 26: 1093–1100.

Lam DSC, Fan DSP, Ng JS, Yu CB, Wong CY

& Cheung AY (2005): Ocular hypertensive

and anti-inflammatory responses to different

dosages of topical dexamethasone in children:

a randomized trial. Clin Experiment Ophthal-

mol 33: 252–258.

Wordinger RJ & Clark AF (1999): Effects of

glucocorticoids on the trabecular meshwork:

towards a better understanding of glaucoma.

Prog Retin Eye Res 18: 629–667.

Received on November 17th, 2009.

Accepted on August 13th, 2010.

Correspondence:

Dorothy SP Fan, FRCS, Department of

Ophthalmology and Visual Sciences,

The Chinese University of Hong Kong, Hong

Kong Eye Hospital, 147K Argyle Street,

Kowloon, Hong Kong, China

Tel: + 852 2762 3150

Fax: + 852 2715 9490

Email: dorothyfan@cuhk.edu.hk

6-month 1-year7-week

Control

DEX

FML

RIM

LOT

(O)(N)(M)

(L)(K)(J)

(I)(H)(G)

(F)(E)(D)

(C)(B)(A)

Fig. 4. Trabecular meshwork region stained by Masson’s trichrome. ECM deposition was more

marked in trabecular meshwork region of eyes treated with dexamethasone (DEX), fluorometho-

lone (FML) and rimexolone (RIM) compared to BSS control. Loteprednol treatment did not

result in increased ECM deposition. Collagen was stained green (with Fast green), plasma was

grey (with acid fuscin and xylidine Ponceau) and cell nuclei were dark brown to black (with Weig-

ert’s iron haematoxylin). (A)–(C) = Control; (D)–(F) = Dexamethasone (DEX); (G)–(I) =

Fluorometholone (FML); (J)–(L) = Rimexolone (RIM); (M)–(O) = Loteprednol etabonate

(LOT).

Acta Ophthalmologica 2010

5