retinal thickening in iridocyclitis

●

ect●

●

tJlnuemnD●

cffm.b(qab0●

cOw(E

Mtmccr

A

CM(

2

0d

Retinal Thickening in Iridocyclitis

CHELSEA G. CASTELLANO, SANDRA S. STINNETT, PRIYATHAM S. METTU, REX M. McCALLUM, AND
GLENN J. JAFFE
wC(

ttejisuetac

tptcebib

C

aoiebwtuntuepahs

w

PURPOSE: To determine the frequency of retinal thick-ning (RT) in eyes with iridocyclitis and to examine theorrelations among anterior chamber (AC) inflamma-ion, RT, and visual acuity.

DESIGN: Retrospective, observational case series.METHODS: Records were reviewed of patients seen at

he Duke University Eye Center Uveitis Clinic fromanuary 2002 through March 2008. Patients with uni-ateral iridocyclitis without panuveitis, vitreitis, pars pla-itis, posterior uveitis, or a combination thereof who hadndergone optical coherence tomography (OCT) of bothyes were included. The AC cell grade and OCT-RTeasurements were recorded. Subretinal fluid and intrareti-al cysts were determined from OCT scans according touke Reading Center guidelines.RESULTS: Forty-three patients were studied. RT typi-

ally was present in a ring-like distribution around theovea. The median difference between the study eye andellow eye in RT was statistically significant for totalacular volume (TMV) and for all OCT subfields (P <

001). In the study eye, there was a modest correlationetween the RT and AC cell grade for the OCT-TMVP � .039; r2 � 0.1) and the subfield comprised of theuadrants in the outer ring on OCT (P � .027; r2 � 0.12),nd between RT and visual acuity for OCT-TMV and allut the central subfields (P � .003 to .007; r2 � 0.261 to.227). RT decreased after anti-inflammatory therapy.CONCLUSIONS: RT is strongly associated with iridocy-

litis and decreases after treatment. RT, as determined byCT, is a useful clinical parameter to evaluate patientsith iridocyclitis and to monitor response to treatment.

Am J Ophthalmol 2009;148:341–349. © 2009 bylsevier Inc. All rights reserved.)

ACULAR EDEMA (ME) IS A KNOWN COMPLICA-

tion of iridocyclitis. Cystoid macular edema(CME) can occur in eyes with a severe inflamma-

ion associated with this condition and also may developore frequently in eyes with nonidiopathic iridocyclitis

ompared with those with idiopathic iridocyclitis.1 Re-ently, it has been suggested that eyes with CME mayepresent the severe end of the spectrum of ME associated

ccepted for publication Mar 23, 2009.From the Department of Ophthalmology, Duke University Medical

enter (C.G.C., S.S.S., P.S.M., G.J.J.); and the Department of Internaledicine, Division of Rheumatology, Duke University Medical Center

R.M.M.), Durham, North Carolina.

cInquiries to Glenn J. Jaffe, Duke Eye Center, Box 3802, Durham, NC

7710; e-mail: [email protected]

© 2009 BY ELSEVIER INC. A002-9394/09/$36.00oi:10.1016/j.ajo.2009.03.034

ith iridocyclitis, but that macular thickening withoutME also can be associated with severe anterior chamber

AC) inflammation.2

Optical coherence tomography (OCT) provides quanti-ative retinal thickness (RT) measurements and qualita-ive morphologic data in eyes with anterior uveitis.3 Forxample, in a cross-sectional case study of patients withuvenile idiopathic uveitis, there was frequent macularnvolvement characterized by perifoveal thickening anderous retinal detachment.4 In a study of patients withnilateral moderate to severe iridocyclitis that excludedyes with mild inflammation, 13 of 15 eyes had diffusehickening on OCT without CME.2 To our knowledge, anssociation between macular thickening and mild iridocy-litis has not been established.

Recently, we have observed RT in eyes with iridocycli-is, even when the AC inflammatory reaction is notronounced. However, the frequency of this occurrence,he relationship between RT and visual acuity (VA), andhanges in RT after anti-inflammatory therapy in theseyes is unknown. Herein, we determined the relationshipetween RT, disease severity, and VA among eyes withridocyclitis. We also determined whether RT was affectedy corticosteroid therapy in a subset of these patients.

METHODS

LINICAL RECORDS OF PATIENTS WITH IRIDOCYCLITIS SEEN

t the Duke University Eye Center Uveitis Clinic by onef us (G.J.J.) between January 2002 and March 2008nitially were identified by analysis of billing codes in anlectronic database. All of the medical records containingilling codes consistent with a diagnosis of iridocyclitisere reviewed individually throughout each patient’s en-

ire follow-up to ensure that iridocyclitis remained the onlyveitis diagnosis during the follow-up. Only patients diag-osed with iridocyclitis were included in this study; pa-ients with panuveitis, vitreitis, pars planitis, posteriorveitis, or a combination thereof were excluded. Otherxclusion criteria included: preexisting macular disease,reexisting ocular or systemic disease that can cause ME,nd prior history of vitreoretinal surgery. Patients with aistory of cataract surgery were included only if they wereeen in the clinic at least 6 months after the operation.

A comprehensive medical and ophthalmologic historyas obtained from each patient during the initial visit. The
linical presentation of individual patients and a detailed
LL RIGHTS RESERVED. 341

mailto:[email protected]

qwpifaoiafllcBarrU

iwamie3acdre

ioSa(sm(msDtcr

Sssmwaaeaam

at

ifican

3

uestionnaire completed by the patient at the first visitith a careful review of systems were reviewed. Allatients underwent a complete ophthalmic examination,ncluding best-corrected VA, biomicroscopy, tonometry,undus examination, and OCT. Fluorescein angiographynd ultrasonography were performed when indicated. Lab-ratory and radiographic investigations were performed asndicated by the clinical presentation or review of systemsnd typically included one or more of the following:uorescent treponemal antibody absorption test for syphi-is, serum angiotensin-converting enzyme, serum antinu-lear antibody, purified protein derivative (PPD), HLA-27 typing, and chest radiograph. Patients with clinicalnd laboratory evidence of associated systemic disease wereeferred to the appropriate medical team, including theheumatologist (R.M.M.) working in conjunction with theveitis Service at Duke.Iridocyclitis was diagnosed based on the presence of

nflammatory cells in the AC and corneal endotheliumith or without flare and absence of posterior vitreous cellsnd other features of posterior segment intraocular inflam-ation other than ME. The degree of anterior segment

nflammation was measured by cell grade in the AC. Eachye received a cell grade of 0, rare, occasional, 1�, 2�,�, or 4� using a grading scale modified from Hogan andssociates.5 Eyes with a cell grade of occasional to 4� wereonsidered to have active inflammation. Eyes previouslyiagnosed with iridocyclitis with an AC cell grade of 0 orare on examination were considered inactive. Mild, mod-

TABLE. Difference in Retinal Thickness Measurements on theMap between the Eye Affected by Irid

Variable Statistic

AC cell grade Mean (SD)

Minimum/median/maximum

Sum of quadrants in inner ring Mean (SD)


Sum of quadrants in outer ring Mean (SD)


Sum of all quadrants in inner

and outer rings

Mean (SD)


Total macular volume Mean (SD)


Center point Mean (SD)


Central 1 mm subfield Mean (SD)


AC � anterior chamber; OCT � optical coherence tomography;aFor each patient, the difference in retinal thickness between the s

for each parameter on the OCT fast macular thickness map: center

of the quadrants in the outer ring, sum of all quadrants in the inn

descriptive statistics were computed for the study eye, for the fellow

was used to test whether the medians of the differences were sign

rate, and severe cases of AC inflammation were included t

AMERICAN JOURNAL OF42

n this study. Iridocyclitis was classified as acute, recurrent,r chronic, as defined by the criteria established by thetandardization of Uveitis Nomenclature Working Group:cute was defined by sudden onset and limited duration�3 months), recurrent was defined by repeated episodeseparated by periods of inactivity without treatment (�3onths), and chronic was defined by persistent uveitis

uveitis lasting � 3 months) with relapse in fewer than 3onths after discontinuing treatment.6 VA with best

pectacle-correction was recorded with Early Treatmentiabetic Retinopathy Study (ETDRS) chart. Logarithm of

he minimum angle of resolution (logMAR) VA wasalculated from the relation: logMAR VA � log of theeciprocal of ETDRS-VA.

Optical coherence tomography was performed using thetratus OCT Fast Macular Thickness Map (software ver-ion 3.0 and 4.0; Zeiss-Meditech, Dublin, California, USA)can mode and a 7-mm line scan centered on the fovea. RTeasurements were obtained by the Stratus machine soft-are algorithm that delineates the inner retinal boundarynd photoreceptor inner segment–outer segment junctionnd determines the distance between these boundaries atach measurement point along the scan’s x-axis and createssurface map reconstruction. Eyes with degraded images

nd images with artifacts sufficient to preclude thicknesseasurements were excluded from this study.The medical treatment of all patients during an acute

ttack of iridocyclitis consisted of cycloplegics and intensiveopical corticosteroids, followed by a slow taper. Some pa-

tus Optical Coherence Tomography Fast Macular Thicknesslitis (study eye) and the Fellow Eyea

Study Eye Fellow Eye Difference P value

1.93 (1.61) 0 (0) 1.93 (1.61)

0/1�/3� 0/0/0 0/1�/3� �.001

194 (192) 1071 (77.3) 123 (195)

5/1165/2142 893/1091/1198 –69/72.5/1157 �.001

028 (105) 933 (78.5) 95.5 (95.1)

2/1013/1262 784/939/1146 –27/58.5/324 �.001

223 (274) 2004 (150) 219 (274)

7/2177/3283 1694/2026/2341 –89/122/1461 �.001

7.55 (0.81) 6.81 (0.53) 0.74 (0.8)

7/7.39/9.92 5.75/6.87/8.13 –0.24/0.54/3.75 �.001

216 (80.2) 174 (29.6) 42.4 (72.2)

2/193/623 119/169/250 –23/18/387 �.001

251 (74.2) 208 (24.9) 42.7 (69.4)

9/236/644 155/212/265 –11/18.5/410 �.001

standard deviation.

ye (affected eye) and the fellow eye (unaffected eye) was computed

central 1 mm subfield, sum of the quadrants of the inner ring, sum

d outer rings, and total macular volume. Then for each variable,

and for the difference between eyes. The Wilcoxon signed-rank test

tly different from 0.

Straocyc

1

89

1

82

2

171

5.9

13

15

SD �

tudy e

point,

er an

eye,

ients required systemic corticosteroids or another immuno-

OPHTHALMOLOGY SEPTEMBER 2009

stWts

gcftgosCoTq

3r6(

●

wetwcTotrs

Ffi

V

uppressive agent, depending on the response to initialherapy, for treatment of the associated systemic disease.

hen ME was present or inflammation was severe or difficulto control, patients additionally were offered periocular orystemic corticosteroids on a case-by-case basis.

Data obtained from the clinical records included: age,ender, race, history of ocular surgery, ocular and systemicomorbidities, duration of iridocyclitis episode, duration ofollow-up, cause and course of the disease, dates and medicalreatments for previous episodes, current treatment, AC cellrade, VA, and intraocular pressure. In addition, the presencer absence of subretinal fluid and intraretinal cysts on OCTcans was determined according to Duke OCT Readingenter guidelines.7 The following RT measurements werebtained and recorded from the Stratus OCT Fast Macularhickness Map: center point, central 1-mm subfield, 4 inner

IGURE 1. Perifoveal retinal thickening in an eye affected byast macular thickness map illustrates diffuse retinal thickeningn the right eye (OD); the left eye (OS) is unaffected.

uadrants that comprise an inner ring that extends from 1 to e

RETINAL THICKENINGOL. 148, NO. 3

mm around the central 1-mm subfield, and 4 outer quad-ants that comprise an outer ring that that extends from 3 to

mm surrounding the inner ring. Total macular volumeTMV) also was recorded.

STATISTICAL METHODS: Only the first visit of patientsith unilateral iridocyclitis who underwent OCT of bothyes was included in the data analysis. For each patient,he difference in RT between the study eye (eye diagnosedith iridocyclitis) and the fellow eye (unaffected eye) wasomputed for each parameter on the OCT Fast Macularhickness Map: center point, central 1-mm subfield, sumf the quadrants in the inner ring, sum of the quadrants inhe outer ring, sum of all quadrants in the inner and outerings, and TMV. Then for each variable, descriptivetatistics were computed for the study eye, for the fellow

cyclitis. This Stratus Optical Coherence Tomography (OCT)) outside the foveal center in a study patient with iridocyclitis

irido(RT

ye, and for the difference between eyes. The Wilcoxon

IN IRIDOCYCLITIS 343

stvbttcam

mWctsmedpunbtbtaasbt

wiUisRTfewC

●

atoqis9pa

f11etsnwtp(dstfT(tempai

●

P

F

r0btsHcd

Ftffb

3

igned-rank test was used to test whether the medians ofhe differences were significantly different from 0. Perifo-eal thickening was defined by a thickness differenceetween the study eye and fellow eye of more than 10% ofhe fellow eye thickness for the sum of the inner ring andhe combined sums of the inner and outer rings. Theut-off value of 10% was chosen because it is thought thatbove this value, the numbers can be distinguished fromeasurement variability.8

Correlation and regression analyses were used to deter-ine the relationship between AC cell grade and RT.hen examining the relationship between VA and RT,

orrelation and regression again were used; however, pa-ients with a visually significant cataract (2� posteriorubcapsular cataract or more or 3� nuclear sclerosis orore) or any other ocular complication affecting VA were

xcluded from the analysis. We also examined the effect ofemographics (age, gender, and race), cause (HLA-B27ositive compared with HLA-B27 negative), and intraoc-lar pressure on RT (while controlling for AC cell grade ifecessary) using correlation and regression analyses orox-and-whisker plots where appropriate. Both the overallhickness in the study eye and the difference in RTetween the study eye and fellow eye were compared withhe various parameters in the correlation and regressionnalyses. For patients with inactive inflammation on ex-mination (cell grade of 0 or rare), correlation and regres-ion analyses were used to determine the relationshipetween the degree of RT and the duration of time fromhe last episode of active AC inflammation.

A separate analysis was performed on a subset of patientsith unilateral iridocyclitis and on 1 patient with bilateral

ridocyclitis who had multiple follow-up visits at the Dukeveitis Clinic to examine the change in RT after anti-

nflammatory therapy. OCT-RT measurements for thetudy eyes of these patients at visit 1 were compared withT measurements at visit 2 using the Student paired t test.he time between visit 1 and visit 2 was variable because

ollow-up visits were not scheduled at the same interval forach patient. All of the statistical analyses were carried outith SAS software version 8.2 (SAS Inc, Cary, Northarolina, USA).

RESULTS

DESCRIPTIVE VARIABLES: A total of 149 patients seent the Duke Uveitis Clinic between 2002 and 2008 methe initial study inclusion and exclusion criteria. However,nly patients with unilateral iridocyclitis who had high-uality OCT scans performed on both eyes were includedn the data analysis, leaving a total of 43 patients in ourtudy population. The mean age was 46 years (range, 7 to2 years); 22 were male and 21 were female. Thirty-fiveatients (81%) were White, 6 patients (14%) were Black,
nd 2 patients (5%) were Asian. The diagnoses were as l

ollows: idiopathic (n � 20), HLA-B27 associated (n �8), herpetic (n � 2), juvenile rheumatoid arthritis (n �), and other (n � 2). Eight patients had acute (firstpisode), 11 had recurrent, and 24 had chronic iridocycli-is. Nine patients (20%) were not receiving any ocular orystemic treatment for iridocyclitis at the time of exami-ation. Twenty-seven patients (64%) were being treatedith topical steroids, and 7 patients (16%) were being

reated with systemic corticosteroids or other immunosup-ressants at the time of their clinic visit. Twenty-four56%) eyes were considered to have active inflammation,efined by an AC cell grade of at least occasional. Ashown in the Table, the median study eye (involved eye)o fellow eye difference in RT was statistically significantor all of the subfields of the OCT Fast Macularhickness Map and for TMV (P � .001). For all eyes

including those study eyes with inactive inflamma-ion), 42% had perifoveal thickening similar to thexample shown in Figure 1. In eyes with active inflam-ation on examination, 63% were found to have

erifoveal thickening. Four patients (9%) had CME,nd 1 patient (2%) had both subretinal fluid and CMEn the study eye.

RELATIONSHIP OF OPTICAL COHERENCE TOMOGRA-

HY RETINAL THICKNESS TO ANTERIOR CHAMBER IN-

LAMMATION: Patients in this study exhibited a broadange of inflammation severity. AC cell grade ranged fromto 3� in the study eye. The degree of difference in RT

etween the study eye and fellow eye for the outer ring ofhe OCT map correlated with the AC cell grade of thetudy eye (P � .027; r2 � 0.12), as shown in Figure 2.owever, the small coefficient of determination (r2) indi-

ates that this was only a modest correlation. Similarly, theegree of difference in TMV between eyes weakly corre-

IGURE 2. Scatterplot showing the difference in RT betweenhe eye affected by iridocyclitis (study eye) and the fellow eyeor the sum of the 4 quadrants of the outer ring on the OCTast macular thickness map compared with the anterior cham-er (AC) cell grade of the study eye (P � .027; r2 � 0.12).

ated with AC cell grade in the study eye (P � .039; r2 �


0c(c

satc..

fet.eibsar

●

P

Tgda

rg

ept.t(mH.tspv

urpgeRtdartdar

●

P

FtffsmF

etmcwo

V

.1). The difference in RT for the inner ring (P � .17),ombined inner and outer rings (P � .079), center pointP � .100), and central 1-mm subfield (P � .13) did notorrelate significantly with AC cell grade.

Similar results were obtained when overall RT in thetudy eye (rather than the difference between study eyend fellow eye) was compared with the AC cell grade inhe study eye. There was a modest correlation between ACell grade to RT in the study eye for the outer ring (P �021; r2 � 0.13), combined inner and outer rings (P �031; r2 � 0.11), and TMV (P � .015; r2 � 0.137).

The mean difference in RT between study eye andellow eye was significantly greater in inflamed studyyes compared with quiet study eyes for all subfields ofhe OCT map and TMV (P values ranged from .004 to025; Figure 3). However, when only actively inflamedyes with AC cell grade greater than occasional werencluded in the analysis (n � 14), the difference in RTetween the study eye and fellow eye did not correlateignificantly with the AC cell grade of the study eye inny of the subfields of the OCT map or TMV (P valuesanged from .26 to .44).


HY RETINAL THICKNESS TO DESCRIPTIVE VARIABLES:

here was no significant correlation between age, race,ender, or IOP and the study eye thickness, or theifference in RT between the study eye and fellow eye for

IGURE 3. Graph showing the difference in RT between theye affected by iridocyclitis (study eye) and the fellow eye forhe sum of the 4 quadrants of the outer ring on the OCT fastacular thickness map for patients with quiet study eyes

ompared with patients with inflamed study eyes. The study eyeas considered to be inflamed if the AC cell grade was equal tor more than occasional cells (P � .004; r2 � 0.19).

ny of the subfields of the OCT map or for TMV (P values m


anged from .061 to .99), while controlling for AC cellrade in subfields when appropriate.

The difference in RT between the study eye and fellowye did not vary by HLA-B27 haplotype (HLA-B27ositive vs HLA-B27 negative) for any of the subfields ofhe OCT map or for TMV (P values ranged from .15 to48). The correlation of HLA-B27 haplotype with RT ofhe study eye was borderline significant for the outer ringP � .054; r2 � 0.17). For all other subfields of the OCTap and for TMV, there was no correlation betweenLA-B27 haplotype and RT (P values ranged from .081 to

53). Additionally, when the relationship of AC cell gradeo the difference in RT between eyes was reexamined aftertratifying the results by HLA-B27 positivity, HLA-B27-ositive status did not seem to affect the correlation (Palues ranged from .36 to .98).The duration of time that study patients were followed

p at the Duke University Eye Center Uveitis Clinicanged from 0 to 120 months, with a mean 11.6 months. Inatients with inactive study eye inflammation (AC cellrade of 0 or rare), the duration of time from the lastpisode of iridocyclitis was compared with the difference inT between the study eye and fellow eye. When time from

he end of the last episode was compared with theifference in RT, no significant correlations were found forny subfields of the OCT map or TMV (n � 11; P valuesanged from .100 to .64). Similarly, when the time fromhe start of the last episode was compared with theifference in RT, no significant correlations were found forny subfields of the OCT map or TMV (n � 15; P valuesanged from .53 to .95).


HY RETINAL THICKNESS TO VISUAL ACUITY: The

IGURE 4. Scatterplot showing the difference in RT betweenhe eye affected by iridocyclitis (study eye) and the fellow eyeor the sum of the 4 quadrants of the outer ring on the OCTast macular thickness map compared to the visual acuity of thetudy eye (P � .003; r2 � 0.26). logMAR � logarithm of theinimum angle of resolution.

ean VA in the study eye was 20/32 (logMAR VA, 0.21)


a(cwe.rd(d((t00cTbw

●

M

iwsttcs(tdDfsptts(q

I

timstdiaRf

tfotDiab(ii2

iclRptilmoTove(tsnwrta

RflacrpfnCuudBoc

the

3

nd VA ranged from 20/20 (logMAR VA, 0) to 20/160logMAR VA, 0.90) for patients without visually signifi-ant cataracts. VA in the study eye modestly correlatedith the difference in RT between the study eye and fellowye for the inner ring (P � .007; r2 � 0.227), outer ring (P �003; r2 � 0.261; Figure 4), and the sum of inner and outerings (P � .004, r2 � 0.255). VA also correlated with theifference in TMV between the study eye and fellow eyeP � .005; r2 � 0.235). VA did not correlate with theifference in RT between eyes for the center point thicknessP � .145; r2 � 0.07) and central 1-mm subfield thicknessP � .080; r2 � 0.100). VA in the study eye correlated withhe RT in the study eye for the inner ring (P � .038; r2 �.14), the sum of the inner and outer rings (P � .042; r2 �.135), center point thickness (P � .013; r2 � 0.19), andentral 1-mm subfield thickness (P � .009; r2 � 0.215).he correlation between VA and RT in the study eye wasorderline significant for TMV (P � .053; r2 � 0.12) andas not significant for the outer ring (P � .070).

RESPONSE OF RETINAL THICKNESS TO TREAT-

ENT: A subset of 9 patients with active unilateralridocyclitis and 1 with active bilateral iridocyclitis, eachith more than 1 follow-up visit and high-quality OCT

cans at all visits, was evaluated to determine the effect ofreatment on RT. All patients in this subset analysis werereated with topical corticosteroids, systemic or periocularorticosteroids, or a combination thereof and demon-trated a reduction in AC cell grade from their first visitvisit 1) to their second visit (visit 2). In this subset, theime between visit 1 and visit 2 was variable. The meanuration of time between visits was 80.9 � 54.7 days.espite the small sample size, the observed decrease in RT

rom visit 1 to visit 2 was statistically significant for theum of the quadrants in the outer ring (P � .016), centeroint thickness (P � .037), and central 1-mm subfieldhickness (P � .049). The difference in thickening be-ween the first 2 visits was borderline significant for theum of the inner and outer rings (P � .056) and TMVP � .057) and was insignificant for the sum of theuadrants in the inner ring (P � .068).

DISCUSSION

N THIS REPORT, WE FOUND A SIGNIFICANT AND CONSIS-

ent correlation between macular thickening and activeridocyclitis. This relationship held, even in eyes with onlyild inflammation. CME is a known complication of

evere iridocyclitis. Traill and associates found that pa-ients with severe iridocyclitis also can exhibit a pattern ofiffuse RT.2 Our findings in patients with less severenflammation are in accordance with those of Traill andssociates in that for patients with unilateral iridocyclitis,T in the affected eye was significantly greater than in the

ellow, unaffected eye. In most eyes with active inflamma- t


ion, there was a ring-like pattern of thickening outside theoveal center in the inner ring and combined inner anduter rings of the affected eye. A similar pattern ofhickening outside of the foveal center was described byucos de Lahitte and associates in a recent study examin-

ng maculopathy in patients with juvenile idiopathicrthritis (JIA) uveitis. This cross-sectional case study ofilaterally-affected JIA uveitis patients found that 4573%) eyes had perifoveolar thickening, defined asncreased thickening in the quadrants of the OCT mapnner ring with an average thickness value of more than88 �m.4

We observed significantly increased RT in eyes withridocyclitis compared with their noninflamed fellow-eyeounterparts. We believe that these differences are un-ikely to be related to physiologic variations. Typically, theT measurements obtained from both normal eyes of a givenatient correlate highly.9,10 Furthermore, the consistent, sta-istically significantly increased thickness measurements innflamed eyes, over a wide range of inflammation severityevels, speaks to a pathophysiologic effect of AC inflam-ation on RT. OCT scans were obtained routinely on eyes

f most patients with iridocyclitis during the study period.herefore, our study population contains a wide spectrumf patients with iridocyclitis, including those with 20/20ision and no clinically evident thickening on clinicalxamination. This practice was initiated when one of usG.J.J.) noted apparent RT in the inflamed eye, relative tohe uninflamed fellow eye. Both eyes were imaged so thatubtle RT in the affected eye could be differentiated fromormal RT, on the assumption that the 2 eyes normallyould have similar thickness when uninflamed. After we

eviewed our data, which confirmed our initial observa-ions, it became our practice to obtain bilateral OCTs inll patients with iridocyclitis.

Whereas most eyes in our study had a pattern of diffuseT, 4 (9%) had CME and 1 had both CME and subretinaluid on OCT. These findings differ from those of Traillnd associates, who did not find any patients with cystoidhanges on OCT.2 Other studies cite an overall occur-ence of CME in 6% to 31% of patients with HLA-B27-ositive anterior uveitis,11–14 with one study citing therequency of CME as 6% in patients with HLA-B27-egative idiopathic anterior uveitis.14 The occurrence ofME has been found in 4% to 9% of patients with acuteveitis and in 3% to 24% of patients with chronicveitis.15,16 In our study, all 5 eyes in which CMEeveloped had chronic iridocyclitis and 2 eyes were HLA-27 positive. These findings, in conjunction with those ofther studies, suggest that the cause and course of iridocy-litis may relate to the development of CME.

The mechanism underlying the development of noncys-ic ME in eyes with iridocyclitis is unknown. However, weypothesize that factors responsible for noncystic ME inyes with iridocyclitis are similar to those that contribute
o uveitic CME. Prostaglandins, inflammatory cytokines,

aitbk

aoallJsloawmptoitpintmtmh

awggAsotfipa

htHffistgfbc

cw

pcwvttmotiHcoFi

wpetHttvtniitutbe

stecOarmlsRdvtotsr

V

nd vascular endothelial growth factor are thought to bemportant contributors to inflammatory CME.11,17,18 It ishought that these factors increase the permeability of thelood-aqueous and blood-retina barriers and promote leu-ocyte migration, contributing to ME formation.We found a modest correlation between the degree of

nterior segment inflammation and the degree of RTutside the foveal center in the outer ring of the OCT mapnd for TMV. De Lahitte and associates found no corre-ation between the degree of inflammation (measured byaser flare cell meter photometry) and RT in patients withIA uveitis (a chronic form of uveitis).4 However, thattudy examined mean foveal thickness only and did notook at the perifoveal areas of the OCT map separately. Inur study, we noted that there was a wide range of RT forgiven cell grade in the study eye, particularly in patientsith a cell grade of occasional or less. This finding in partay be because of the inclusion of a large number of

atients with chronic iridocyclitis whose affected eyesended to have thick retinas despite a relatively low gradef anterior segment inflammation. Patients with chronicridocyclitis may have persistently low levels of inflamma-ory mediators present for longer periods of time, leading torolonged periods of ME compared with those with moresolated episodes of iridocyclitis. Another possible expla-ation for the wide range of RT for a given cell grade ishat there may be a lag between inflammation and ME. RTay reflect inflammation activity in the past, indicating

hat our patients with minimal or no signs of AC inflam-ation on examination but evidence of RT on OCT mightave been recovering from a recent episode of iridocyclitis.When the mean difference in RT between the study eye

nd fellow eye was examined in quiet study eyes comparedith actively inflamed study eyes, inflamed eyes had areater mean thickness in all subfields of the OCT map andreater TMV. Given the widely variable RT in eyes withC cell grades of occasional or less, we expected to find a

tronger correlation between RT and AC cell grade whennly the actively inflamed eyes with a cell grade of morehan occasional were examined. Surprisingly, we did notnd correlations on any of the subfields of the OCT map,erhaps reflecting the small sample size in this subsetnalysis.

Correlations of demographics with macular thickeningave been examined in healthy eyes. There does not seemo be a relationship between age and foveal thickness.19

owever, Kelty and associates recently found that theovea is significantly less thick in Black persons andemales than in White persons and males.9 Our resultsndicated that age, gender, race, and IOP do not have aignificant effect on RT in patients with iridocyclitis andhat these factors cannot account for the effect of AC cellrade on the difference in RT between the study eye andellow eye. Additionally, we found no difference in RTased on HLA-B27 haplotype. Other studies produced
onflicting results regarding the complications and out- u

omes associated with HLA-B27-positive status in patientsith anterior uveitis.13,14,20,21

Macular edema is a significant cause of vision loss inatients with uveitis. A previous study found a negativeorrelation between VA and macular thickness in patientsith uveitic ME.22 De Lahitte and associates found thatisual impairment increased with foveal thickness in pa-ients with JIA.4 Previous studies focused mainly on fovealhickness in relationship to VA. Our results indicate aodest correlation between VA and RT in the subfields

utside the foveal center on the OCT map, indicating thathe diffuse RT in patients with iridocyclitis that occurredn the perifoveal area may have an impact on VA.owever, the observed variable RT measurements that

orresponded to a given VA level indicate that RT is justne of many variables affecting vision in these patients.uture studies are needed to evaluate further the clinicalmportance of RT in patients with iridocyclitis.

When the time from the last episode of inflammationas compared with the difference in RT between eyes inatients with inactive inflammation on examination, wexpected to see decreased average RT as the amount ofime elapsed from the last episode of iridocyclitis increased.owever, we did not find a significant correlation between

he difference in RT between eyes and the interval fromhe last episode. In our study, follow-up intervals wereariable and the dates of patients’ most recent inflamma-ory episodes were obtained from patient history and wereot always known precisely. Some patients with inactive

nflammation on examination were excluded because thisnformation was not recorded, leaving only 11 patients inhis subgroup analysis. These limitations prevented us fromsing defined time points and might have contributed tohe lack of association between the difference in RTetween eyes and the duration of time from the lastpisode of inflammation.

When we examined the change in RT as anterioregment inflammation resolved in the study eyes of pa-ients with follow-up visits, there was a significant differ-nce in RT from the first visit to the second visit for theenter point, central 1-mm subfield, and outer ring of theCT map, despite the small sample size in the subset

nalysis. The observed decrease in RT as inflammationesolved suggests that RT measurements can be useful toonitor the therapeutic response over time. Because of the

imited number of patients with more than 2 visits in ourtudy, we were unable to assess adequately the resolution ofT statistically over longer periods of follow-up or toetermine any differences in the change in RT amongarious subgroups of patients, such as those treated solely withopical steroids compared with those treated with oral steroidsr steroid injections. Interestingly, Traill and associates foundhat 5 of 11 patients available for follow-up still showedignificant macular thickening at 6 months despite theesolution of clinical signs and symptoms of acute anterior
veitis.2 It is possible that there is variability in the degree

opota

phimcfmpRpmui

chhhso

csooosrafaa

TTCaa

3

f thickness resolution after episodes of iridocyclitis, de-ending on the cause and course of the disease, as well asther clinical and genetic factors. Further investigationshat include larger numbers of patients will be necessary toddress these questions.

This retrospective study has limitations. Many of theatients were referred to our uveitis clinic after they firstad been seen by an outside referring physician who

nitiated therapy at the time of presentation. This factight have caused a referral bias toward more severe and

omplex disease. Furthermore, this limitation prevented usrom seeing many patients at the onset of their disease andight have accounted for unresolved thickening from

revious inflammatory episodes and the variable degrees ofT seen when we first evaluated the patient. Someatients with very severe inflammation with associatededia opacity were excluded from our study because of

nreadable OCT scans. Had we been able to measure RT
n these eyes, we might have been able to detect a stronger u
BM. Macular thickness assessment in healthy eyes based on

1

1

1

1

1

1

1

1

1


orrelation between RT and degree of inflammation, or aigher proportion of eyes with CME. It also is possible thatigher-resolution spectral-domain OCT machines mightave identified an even greater proportion of eyes withubtle cystic changes; such a study currently is underway atur institution.The strong and consistent correlation between in-

reased RT and iridocyclitis observed in the presenttudy suggest an important, clinically measurable effectf anterior uveitis on RT. Additionally, the correlationf RT with the degree of AC inflammation and thebserved reduction in RT as inflammation resolveduggest that RT may be a useful parameter to monitoresponse to therapy. A prospective study to evaluate RTt baseline and after treatment is needed to examineurther the resolution of RT over time and wouldddress definitively the usefulness of RT as a marker tossess response to therapy. Such a study currently is
nderway at our institution.
HE AUTHORS INDICATE NO FINANCIAL SUPPORT. DR STINNETT HAS A CONSULTING RELATIONSHIP WITH SIRIONherapeutics. Involved in design and conduct of study (G.J.J., C.G.C., R.M.M.); collection, management, analysis, and interpretation of data (P.S.M.,.G.C., S.S.S., G.J.J.); and preparation, review, or approval of the manuscript (G.J.J., C.G.C., P.S.M., S.S.S., R.M.M.). Duke Institutional Review Board

pproval was obtained to conduct this study and waived consent and Health Insurance Portability and Accountability Act authorization. The studydheres to the Declaration of Helsinki.

REFERENCES

1. Menezo V, Lightman S. The development of complicationsin patients with chronic anterior uveitis. Am J Ophthalmol2005;139:988–992.

2. Traill A, Stawell R, Hall A, Zamir E. Macular thickening inacute anterior uveitis. Ophthalmology 2007;114:402.

3. Antcliff RJ, Stanford MR, Chauhan DS, et al. Comparisonbetween optical coherence tomography and fundus fluores-cein angiography for the detection of cystoid macular edemain patients with uveitis. Ophthalmology 2000;107:593–599.

4. de Lahitte GD, Terrada C, Tran TH, et al. Maculopathy inuveitis of juvenile idiopathic arthritis: an optical coherencetomography study. Br J Ophthalmol 2008;92:64–69.

5. Hogan MJ, Kimura SJ, Thygeson P. Signs and symptoms ofuveitis. I. Anterior uveitis. Am J Ophthalmol 1959;47:155–170.

6. Jabs DA, Nussenblatt RB, Rosenbaum JT, Standardization ofUveitis Nomenclature (SUN) Working Group. Standardizationof uveitis nomenclature for reporting clinical data. Results ofthe First International Workshop. Am J Ophthalmol 2005;140:509–516.

7. Zhang N, Hoffmeyer GC, Young ES, et al. Optical coherencetomography reader agreement in neovascular age-related mac-ular degeneration. Am J Ophthalmol 2007;144:37–44.

8. Diabetic Retinopathy Clinical Research Network, KrzystolikMG, Strauber SF, Aiello LP, et al. Reproducibility of macularthickness and volume using Zeiss optical coherence tomog-raphy in patients with diabetic macular edema. Ophthalmol-ogy 2007;114:1520–1525.

9. Kelty PJ, Payne JF, Trivedi RH, Kelty J, Bowie EM, Burger

ethnicity using Stratus OCT optical coherence tomography.Invest Ophthalmol Vis Sci 2008;49:2668–2672.

0. Bressler NM, Edwards AR, Antoszyk AN, et al, DiabeticRetinopathy Clinical Research Network. Retinal thicknesson Stratus optical coherence tomography in people withdiabetes and minimal or no diabetic retinopathy. Am JOphthalmol 2008;145:894–901.

1. Uy HS, Christen WG, Foster CS. HLA-B27-associateduveitis and cystoid macular edema. Ocul Immunol Inflamm2001;9:177–183.

2. Dodds EM, Lowder CY, Meisler DM. Posterior segment inflam-mation in HLA-B27 � acute anterior uveitis: clinical charac-teristics. Ocul Immunol Inflamm 1999;7:85–92.

3. Chang JH, McCluskey PJ, Wakefield D. Acute anterior uveitisand HLA-B27. Surv Ophthalmol 2005;50:364–388.

4. Power WJ, Rodriguez A, Pedroza-Seres M, Foster SC. Out-comes in anterior uveitis associated with the HLA-B27haplotype. Ophthalmology 1998;105:1646–1651.

5. Rothova A, van Veenedaal WG, Linssen A, Glasius E,Kijlstra A, de Jong PT. Clinical features of acute anterioruveitis. Am J Ophthalmol 1987;103:137–145.

6. Tay-Kearney ML, Schwam BL, Lowder C, et al. Clinicalfeatures and associated systemic diseases of HLA-B27 uveitis.Am J Ophthalmol 1996;121:47–56.

7. Guex-Crosier Y. The pathogenesis and clinical presentationof macular edema in inflammatory diseases. Doc Ophthalmol1999;97:297–309.

8. Luna JD, Chan CC, Derevjanik NL, et al. Blood-retinal(BRB) breakdown in experimental autoimmune uveoretini-tis: comparison with vascular endothelial growth factor,tumor necrosis factor alpha, and interleukin-1beta-mediated
breakdown. J Neurosci Res 1997;49:268–280.

1

2

2

2

V

9. Chan A, Duker JS, Ko TH, Fujimoto JG, Schuman JS.Normal macular thickness measurements in healthy eyesusing Stratus optical coherence tomography. Arch Ophthal-mol 2006;124:193–198.

0. Wakefield D, Easter J, Robinson P, Penny R. Immunologicalfeatures of HLA-B27 anterior uveitis. Aust J Ophthalmol
1983;11:15–19.

1. Linssen A, Meenken C. Outcomes of HLA-B27-positive andHLA-B27-negative acute anterior uveitis. Am J Ophthalmol1995;120:351–361.

2. Markomichelakis NN, Halkiadakis I, Pantelia E, et al.Patterns of macular edema in patients with uveitis: qualita-tive and quantitative assessment using optical coherence
tomography. Ophthalmology 2004;111;946–953.

CHr

3

Biosketch

helsea Castellano will receive an MD from Duke University School of Medicine, Durham, North Carolina in May 2009.er research endeavors were mentored by Dr Glenn J. Jaffe at the Duke University Eye Center. She hopes to pursue

esidency training in Ophthalmology in an academic setting that combines clinical training and research.

AMERICAN JOURNAL OF OPHTHALMOLOGY49.e1 SEPTEMBER 2009

retinal thickening in iridocyclitis

Documents