Incidence and Progression of Epiretinal Membranes inEyes After Cataract Surgery

CALVIN SZE-UN FONG, PAUL MITCHELL, ELENA ROCHTCHINA, THOMAS HONG, TANIA DE LORYN,AND JIE JIN WANG

� PURPOSE: To assess eye-specific epiretinal membrane(ERM) incidence 3 years after phacoemulsificationsurgery, and ERM detection bias attributable to cataract.� DESIGN: Cohort study.� METHODS: We recruited 1932 cataract surgicalpatients aged ‡64 years at Westmead Hospital (2004-2007). The surgical eye of each patient was assessed forpresence of cellophane reflex or preretinal fibrosis atpreoperative and 1-month-postoperative visits, and annu-ally thereafter, using retinal image grading. Agreement onERM detection between preoperative and 1-month-postoperative visits was assessed using kappa statistics.Cumulative incidence of ERM from 1 month to 3 yearspostoperatively was estimated using Kaplan-Meiermethods and compared to the 5-year incidence of idio-pathic ERM in right eyes of age-matched Blue MountainsEye Study (BMES) participants.� RESULTS: ERM prevalence was 13.9% among 1394participants with retinal photographs taken 1 monthpostoperatively. Of 1040 participants with retinalphotographs from both preoperative and 1-month-postoperative visits, ERM was detected in 3.1% and14.8%, respectively, with low diagnostic agreement(kappa[ 0.17). Of 1119 subjects without ERM 1monthpost surgery, the 3-year cumulative incidence of ERMwas 11.2% (95% confidence interval [CI], 9.4%-13.4%; cellophane reflex 6.6%; preretinal fibrosis4.2%). The age-standardized 3-year incidence of ERMin the surgical cohort (12.1%, 95% CI 8.6%-16.9%)was higher than the 5-year incidence of the BMESsubsample (4.4%, 95% CI 3.0%-6.0%).� CONCLUSIONS: A substantial under-detection of ERMin eyes before cataract surgery could incorrectlycontribute to ERM incidence after surgery. Over 3 years,ERM developed in >10%, including preretinal fibrosisin 4%, of surgical eyes free of ERM 1 month postsurgery. (Am J Ophthalmol 2013;156:312–318.� 2013 by Elsevier Inc. All rights reserved.)

Accepted for publication Mar 18, 2013.From the Centre for Vision Research, Department of Ophthalmology

and Westmead Millennium Institute, University of Sydney, Sydney,Australia.

Inquiries to Professor Jie Jin Wang, Centre for Vision Research,Department of Ophthalmology, University of Sydney, WestmeadHospital, Hawkesbury Road, Westmead, NSW Australia, 2145; e-mail:jiejin.wang@sydney.edu.au

312 � 2013 BY ELSEVIER INC.

EPIRETINALMEMBRANES (ERM) ARE A FREQUENT, AGE-

related eye condition, occurring in 6%-19% of eyesin persons aged 40 years or older.1–3 The ERM

prevalence in post–cataract surgical eyes ranged from17%-28% in previous population-based studies.1,3,4 Onfundus photographs, ERM present as 2 forms: cellophanereflex, an early, usually asymptomatic form; andpreretinal fibrosis, a more severe form with associatedfolds. ERM have been shown in previous observationalstudies to be associated with past cataract surgery.4–6 Ascataract may obscure a clear view of fundus photographs,4

ERM prevalence could be markedly underestimated inpreoperative eyes and the incidence of new ERM overesti-mated postoperatively. To assess incident ERM in eyesafter cataract surgery, it is necessary to separate the contri-bution of increased detection from a true increase in thedevelopment of ERM in these eyes.In this study, we aimed to assess detection bias by exam-

ining the difference in ERM detection in the same eyesbefore and after cataract surgery using retinal images takenbefore and soon after cataract surgery. We also aimed toassess the incidence and progression of ERM overa 3-year period from 1-36 months after phacoemulsificationcataract surgery in an older cataract surgical cohort, theAustralian Prospective Cataract Surgery and Age-relatedMacular Degeneration Study. Lastly, we aimed to compareERM incidence between this cohort (eyes after cataractsurgery) and a subsample of the Blue Mountains Eye Study(BMES), who were aged >_64 years and had not had cataractsurgery.

METHODS

THE AUSTRALIAN PROSPECTIVE CATARACT SURGERY AND

Age-related Macular Degeneration Study is a hospital-based, prospective cohort study of cataract surgical patientsaged >_64 years recruited between 2004 and 2007 fromWestmead Hospital, a tertiary referral center, and privateophthalmology clinics in urban districts of western Sydney,Australia. The study adhered to the tenets of the Declara-tion of Helsinki and was approved by the Human ResearchEthics Committees of the University of Sydney and theSydney West Area Health Service. All patients providedwritten informed consent. Details of the study have beenpreviously described.7

0002-9394/$36.00http://dx.doi.org/10.1016/j.ajo.2013.03.022

ALL RIGHTS RESERVED.

Of 1932 patients recruited for phacoemulsificationsurgery at Westmead Hospital, 1861 patients had under-gone surgery, attended 1-month-postoperative visits, andwere then followed annually up to 36 months after theirsurgery. All patients underwent phacoemulsificationsurgery with capsulorrhexis (mean diameter 5.2 mm),hydrodissection, and intraocular lens (IOL) implantation.Of the 71 patients who did not attend 1-month-postopera-tive visits, 63 either had not yet had cataract surgery or hadsurgery performed elsewhere, 7 withdrew consent beforethe 1-month visit, and 1 had an undetermined type ofeye surgery; thus, they were excluded.

In this report, the following patients were excluded: 110with preoperative ocular conditions known to predispose toERM, including moderate-to-severe diabetic retinopathy(modified Airlie House, Early Treatment Diabetic Reti-nopathy Study [ETDRS] score >_43), retinal vascular occlu-sion, and cystoid macular edema; and 256 who did notattend any of the postoperative visits at either 12, 24, or36 months, which included those who died (n ¼ 72),were too ill (n ¼ 43), refused (n ¼ 101), lived too faraway (n ¼ 17), or were uncontactable (n ¼ 23).

Of the remaining 1495 patients, 1394 (93.2%) hadretinal images of sufficient quality for analysis of ERM prev-alence. Of the 101 (6.8% of 1495) patients who did nothave retinal images taken at the 1-month visit, 21(20.8% of 101) refused photography, and the remainderhad no specific reason for the lack of 1-month images.

In order to compare ERM detected preoperatively and1 month postoperatively in the same eyes, 1040 of the 1394patients had available images from both visits and thus wereincluded in this analysis. Of the 1394 patients, 193 hadERM at 1 month, and 82 had ungradable retinal images atfollow-up visits, leaving 1119 patients who had no evidenceof ERM at the 1-month visit but had gradable images at 1or more follow-up visits to determine ERM incidence.

The BMES is a population-based prospective cohortstudy of a suburban Australian population aged 49 yearsor older at baseline. This study also adhered to the tenetsof the Declaration of Helsinki and was approved by theHuman Research Ethics Committees of the University ofSydney and the Sydney West Area Health Service. Allparticipants provided written informed consent. Surveymethods and procedures, including stereo retinal photog-raphy, were described elsewhere.1,8 In summary, of the3654 baseline (1992-1994) participants, 2334 (75.0% of3111 survivors) were reexamined at 5-year follow-up visits(1997-1999). For comparison of ERM incidence in theBMES population with that in the Australian ProspectiveCataract Surgery and Age-related Macular DegenerationStudy cohort, we selected a subsample of BMES partici-pants aged >_64 years at baseline who had no ocular pathol-ogies (moderate-to-severe diabetic retinopathy, retinalvascular occlusion, or cystoid macular edema) that predis-pose to ERM formation, substantial cataract, or cataractsurgery (n ¼ 781, 33.5% of 2334). Substantial cataract

VOL. 156, NO. 2 EPIRETINAL MEMBRANES IN EYES

was defined as presence of nuclear cataract >_ standardphotograph 4 of the Wisconsin cataract grading system,9

posterior subcapsular cataract graded as >_5% of the lensarea, or cortical opacity >_25% of the lens area.10

� MEASUREMENTS: Eye examinations were conductedpreoperatively and at 1, 12, 24, and 36 months postopera-tively. Presenting visual acuity (VA), with habitual correc-tion if worn, and pinhole-corrected VA were measuredusing a back-illuminated logarithm of the minimal angleof resolution (logMAR; Vectorvision CSV 1000, Vectorvi-sion Inc, Dayton, Ohio, USA) chart.Mydriatic retinal photography of both eyes, including

ETDRS standard field 1 (optic disc) and field 2 (macula),11

was performed at each visit, using a retinal camera (TopconTRC 50 IA; Topcon Optical, Tokyo, Japan) with Kodach-rome 64 35-mm slide film or a mydriatic digital camera(Canon CF-60DSi; Canon, Tokyo, Japan). The classifica-tion of ERM lesions was similar to that used in theBMES,1 which was adopted from the Beaver Dam EyeStudy (BDES).4 ERM were classified into 2 forms: an earlyand less severe form known as cellophane reflex; and a laterform, preretinal fibrosis, that included retinal folds. Eyeswith both cellophane reflex and preretinal fibrosis wereclassified as preretinal fibrosis only. Grading of preoperativephotographs and side-by-side grading of the 1-, 12-, 24-,and 36-month photographs taken at the postoperativevisits were performed by an experienced grader, with adju-dication by a senior researcher (J.J.W.) and a retinalspecialist (P.M.).Incidence of ERM was determined if either cellophane

reflex or preretinal fibrosis was newly found at any of theannual follow-up visits in operated eyes that had no signsof ERM at baseline (Figures 1 and 2). Progression of ERMwas defined when the area of preretinal fibrosis orcellophane reflex was detected to increase by >25% of theinitial area from the 1-month visit to any subsequent post-operative visits (Figure 3). Regression of ERM was definedby any of the following criteria: (1) decrease by 25% ormore in the area involved; (2) disappearance of ERM; or(3) preretinal fibrosis reverting to cellophane reflex.Demographic data and a medical history were collected

preoperatively and verified against patient medical records.

� STATISTICALANALYSIS: SAS v 9.13 (SAS Institute Inc,Cary, North Carolina, USA) was used for statistical anal-yses. The agreement in ERM detection between the preop-erative and 1-month-postoperative visits was tested usingsimple kappa coefficients. Both the prevalence and inci-dence of ERM were defined using the study (operated)eye of each patient, and only 1 eye of each patient wasincluded in the analyses. For patients who had alreadyhad cataract surgery in their first eye before study recruit-ment, the study eye was the second eye that was operatedon after recruitment. The cumulative incidence rates forERM and their subtypes, cellophane reflex and preretinal

313AFTER CATARACT SURGERY

FIGURE 1. Development of cellophane reflex in the right eye of a 76-year-old man (pinhole visual acuity 6/9 from 1-36 months).(Left) Normal macula at the 1-month-postoperative examination. (Right) After 36 months, cellophane reflex mainly nasal to themacula is apparent.

FIGURE 2. Development of preretinal fibrosis in the left eye of a 66-year-old man (pinhole visual acuity 6/6 from 1-24 months).(Left) Retina with a few macular soft drusen at the 1-month-postoperative examination. (Right) After 24 months, preretinal fibrosiswith a puckered appearance is apparent.

fibrosis, from the postoperative 1-month to 36-monthperiod were calculated using Kaplan-Meier (product-limit)estimates. The 36-month cumulative incidence of ERM ineyes after cataract surgery in the Australian ProspectiveCataract Surgery and Age-related Macular DegenerationStudy was retrospectively compared to the 5-year idio-pathic ERM incidence in the right eyes of BMES partici-pants aged 64þ years. Ninety-five percent confidenceintervals (CI) were reported.

RESULTS

THE MEAN AGE OF THE 1394 PATIENTS WHO ATTENDED

1-month-postoperative visits was 74.9 years (standard devi-ation 5.9 years). Of the 1394 patients, 87.1% (n ¼ 1214)and 77.4% (n ¼ 1079) attended the 24- and 36-monthvisits, respectively. Apart from age, the preoperative char-acteristics of the 1394 patients who attended were similarto those of participants who had withdrawn (Table 1).

Of 1394 surgical eyes with 1-month-postoperativeimages, ERM prevalence was 13.8% (n¼ 193) (cellophanereflex 7.8%; preretinal fibrosis 6.0%). Of 1040 eyes withboth preoperative and 1-month-postoperative images,

314 AMERICAN JOURNAL OF

3.1% (n ¼ 32) and 14.8% (n ¼ 154) were found to haveERM at the preoperative and 1-month-postoperative visits,respectively. Of the 154 eyes with ERM detected at the1-month-postoperative visit, 87.0% (n ¼ 134) were notdetected preoperatively (Table 2). Kappa was 0.17between ERM detection at the preoperative and 1-month-postoperative visits of the same eyes. Of 146 (14.0% ofthe 1040) eyes that were classified differently betweenthe preoperative and 1-month-postoperative visits, 65.1%(n ¼ 95) had macula-centered images with poor focusfrom the preoperative visit, while 15.8% (n ¼ 23) hadpoor-quality images from the 1-month-postoperative visit.From the 1- to 36-month-postoperative visits, 6.8%

(n ¼ 82) of 1201 patients without ERM at the 1-monthvisit had insufficient follow-up data, leaving 1119 patientsfor analysis of ERM incidence. Table 3 shows that thecumulative incidences for cellophane reflex, preretinalfibrosis, and overall ERM 36 months after cataract surgerywere 6.6% (95% CI 5.3%-8.3%), 4.2% (95% CI 3.1%-5.6%), and 11.2% (95% CI 9.4%-13.4%), respectively.Of the 1119 study eyes, 17.4% (n = 195) had poor-quality images from the 1-month visits and 11.8% (n =132) from the 36-month follow-up visits.The proportion of patients with preretinal fibrosis at the

36-month-postoperative visit who had pinhole VA <6/12

AUGUST 2013OPHTHALMOLOGY

TABLE 2.Comparison of Epiretinal Membrane Prevalence atPreoperative and 1-Month-Postoperative Visits in the

Australian Prospective Cataract Surgery and Age-related

Macular Degeneration Study

1-Month-Postoperative

Participants, N (%)

ERM None Total

Preoperative

participants

ERM a b

20 (13.0) 12 (1.3a) 32 (3.1)

None c d

134 (87.0b) 874 (98.7) 1008 (96.9)

Total 154 (100) 886 (100)

Agreement ¼ 86.0%

Simple kappa ¼ 0.17

(95% CI 0.10-0.25)

1040 (100)

CI ¼ confidence interval; ERM ¼ epiretinal membranes.aFalse positive rate (b / b þ d).bFalse negative rate (c / a þ c).

TABLE 1. Comparison of Baseline Characteristics ofParticipants who Attended Any Postoperative Visit After the

1-Month Post–Cataract Surgical Follow-up and Participants

who Withdrew After 1-Month Follow-up Visit

Characteristic

Included

Participants

(N ¼ 1394)

Withdrawn

Participants

(N ¼ 255) P Value

Age, mean years 6 SD 74.9 6 5.9 76.4 6 6.8 .0013

Female, % 57.7 61.6 .25

Diabetes, % 24.6 28.6 .17

Coronary artery disease, % 22.3 27.5 .070

Prior cataract surgery, % 25.3 26.5 .69

AMD, % 16.5 13.5 .26

AMD ¼ age-related macular degeneration.

FIGURE 3. Development of cellophane reflex and subsequent progression to preretinal fibrosis in the right eye of a 69-year-old man(pinhole visual acuity 6/9 at 1, 24, and 36 months). (Left) Retina with early age-related macular degeneration at the 1-month-postoperative examination. (Center) After 24 months, cellophane reflex inferior to the macula. (Right) After 36 months, extensivecellophane reflex with a focus of preretinal fibrosis temporally is evident.

was 14% (n ¼ 6 of 43). This proportion was not signifi-cantly different from the corresponding proportions withthe same level of VA among eyes with either cellophanereflex only (8%, n ¼ 6 of 71; P ¼ .35) or no ERM at all(14%, n ¼ 139 of 1004; P ¼ .98). One of the 1119 partic-ipants did not have VA data and was excluded from theVA analysis. The proportion of patients who had lost >_2logMAR lines of vision from 1 to 36 months were similarin the groups with preretinal fibrosis (33%, n ¼ 14 of 43),cellophane reflex (21%, n ¼ 15 of 71), and no ERM(32%, n¼ 321 of 1004) (P¼ .46 for each group comparisonwith the preretinal fibrosis group).

Table 4 shows the progression and regression of ERMover the 3 years after cataract surgery. Of 133 eyes withcellophane reflex at the 1-month examinations, 45.1%(n ¼ 60) progressed, including 13.5% (n ¼ 18) thatprogressed to preretinal fibrosis. Over the same period oftime, 27.8% (n ¼ 37) of 133 cellophane reflex cases and18.9% (n ¼ 14) of 74 preretinal fibrosis cases detected1 month postoperatively had regressed spontaneouslywithout surgical intervention (membrane peeling). Of the14 cases of preretinal fibrosis regression, only 5 hadcomplete regression, while the rest had partial regression

VOL. 156, NO. 2 EPIRETINAL MEMBRANES IN EYES

of ERM. After excluding 57 subjects who had poor-qualityimages at 1 or more follow-up visits, the regression rateswere similar: 26% (n ¼ 24 of 93) of cellophane reflex casesand 18% (n¼ 10 of 57) of preretinal fibrosis cases regressed.The 36-month age-standardized ERM incidence after cata-

ract surgery was significantly higher in this surgical cohort(12.1%, 95% CI 8.6%-16.9%) than the 5-year incidence ofidiopathicERMin theBMESsubsampleof the sameage range(4.4%, 95%CI 3.0%-6.0%), and this increased incidencewasconsistently observed across all age groups (Table 5).

DISCUSSION

PREVALENCE RATES FOR THE EARLY (CELLOPHANE REFLEX)

and more severe forms of ERM (preretinal fibrosis) in

315AFTER CATARACT SURGERY

TABLE 3. 12- and 36-Month Cumulative Incidence Rates ofEpiretinal Membranes in the Australian Prospective Cataract

Surgery and Age-related Macular Degeneration Study,

Among 1119 Eyes Without Epiretinal Membranes at the1-Month-Postoperative Visit

Cumulative Incidence,

% (95% CI)

Category of ERM

Cellophane

Reflex

Preretinal

Fibrosis Total ERM

12-month 3.4 (2.5-4.6) 1.6 (1.0-2.5) 5.0 (3.9-6.5)

24-month 5.8 (4.5-7.3) 3.4 (2.4-4.7) 9.0 (7.5-10.9)

36-month 7.0 (5.6-8.8) 4.4 (3.3-5.9) 11.2 (9.4-13.4)

CI ¼ confidence interval; ERM ¼ epiretinal membranes.

TABLE 4. Progression or Regression of Pre-existingEpiretinal Membranes in Affected Cases From 1-36 Months

Postoperatively

3-Year Change, N (%)

Type of ERM After 1 Month

Cellophane

Reflex (N ¼ 133)

Preretinal

Fibrosis (N ¼ 74) ERM (N ¼ 207)

Progressed 60 (45.1) 29 (39.0) 89 (43.0)

Stable 36 (27.1) 31 (42.0) 67 (32.4)

Regressed 37 (27.8) 14 (19.0) 51 (24.6)

ERM ¼ epiretinal membranes.

Australian Prospective Cataract Surgery and Age-RelatedMacular Degeneration Study patients 1 month after surgerywere 8% and 6%, respectively. Three-year cumulative inci-dence rates for these lesions were 7% and 4%, respectively.We found poor diagnostic agreement on ERM detectionbetween grading of preoperative and postoperative imagesof the same eyes by the same photographic grader. Over13% of eyes with cellophane reflex progressed to preretinalfibrosis and 28% regressed spontaneously 3 years after cata-ract surgery, while 19% of eyes with preretinal fibrosisregressed spontaneously, often partially, over the sameperiod of time. Supplemental analysis after exclusion ofsubjects with poor-quality images confirmed that theobserver regression rates of these lesions was unlikelyattributable to the detection issue.

In our study, 87% of eyes that were assessed as havingERM 1 month after cataract surgery were missed in theassessment of their preoperative retinal images. Two thirdsof preoperative images were poorly focused, because ofeither media opacity or photographic technique, whichcontributed to the substantial under-detection of ERM.

Previous population-based studies that reported an asso-ciation between cataract surgery and ERM did not specifythe type of cataract surgery performed.1,4 In the pastdecade, phacoemulsification surgery has become the mostfrequent technique for cataract surgery in the developedworld.12 Previous authors have hypothesized that the pres-ence of cataract may result in under-detection of ERM.6,13

Jahn and associates assessed this question in a clinical studyand found that the eye-specific prevalence of ERM 2 weeksafter phacoemulsification surgery was significantly higher(15%) than the prevalence estimated 2 weeks prior tosurgery (6%).6 Our study confirmed the findings by Jahnand associates and showed that 87% of eyes with ERMdetected shortly after cataract surgery were not detectedpreoperatively. In keeping with the findings of Jahn andassociates and our findings, another prospective study of45 cataract surgical patients, using optical coherencetomography (OCT) to detect ERM, did not find any newcases of ERM between preoperative (unspecified date)and 6-week-postoperative visits.14

316 AMERICAN JOURNAL OF

In comparison with the 5-year incidence of idiopathicERM among BMES participants of similar ages who hadno cataract surgery and no known secondary cause forERM (4%, 95% CI 3%-6%), our surgical cohort hadsubstantially higher incidence of ERMover the 3-year-post-operative period (12%, 95%CI 9%-17%) (Table 5). Giventhat some BMES participants could have had cataract thataffected detection of ERM, we excluded participants whohad substantial cataract from the subsample. In addition,most people in the general older population with a mildlevel of cataract that does not require surgery would havehad sufficient view of the retina to detect ERM. For thisreason, under-detection of ERM incidence in the BMESpopulation, if present, would be limited in magnitude.The strengths of our study included its prospective

cohort design, large sample size, the availability of imagesfrom eyes 1 month postoperatively to estimate ERM prev-alence without the effect of cataract (thus enabling exclu-sion of the prevalent cases for incidence assessment), andthe detailed grading of ERM using standardized methods.In addition, we used preoperative and immediately postop-erative images from the same eyes to document substantialunder-detection of ERM in eyes with cataract prior tosurgery, which could have partly contributed to theincreased ERM prevalence4,13 or incidence15 in post–cataract surgical eyes reported previously.An important limitation of our study is the unavail-

ability of OCT imaging, and a secondary limitation is themoderate dropout rate (15%). OCT could have been usefulto detect ERM detachment from the posterior vitreouscortex. Apart from the slightly younger mean age ofpatients who were followed compared to those who hadwithdrawn, the 2 groups were similar in other preoperativecharacteristics (Table 1). Side-by-side grading of retinalimages from the same visit and multiple visits of our studywere performed by a single grader who has extensive expe-rience in photographic grading of various retinal condi-tions and lesions, having graded all images from allBMES visits for these lesions.16–18 It was possible thatunder- or overestimation of ERM incidence could haveoccurred, but this was not likely to be substantial, ifpresent, as most subjects had multiple retinal images

AUGUST 2013OPHTHALMOLOGY

TABLE 5. Comparison of the Cumulative Incidence of Epiretinal Membranes Between the 3-Year Follow-up Examination AfterCataract Surgery in the Australian Prospective Cataract Surgery and Age-related Macular Degeneration Study and the 5-Year

Follow-up Examination of the Blue Mountains Eye Study

Baseline Age, y

BMESa Australian Prospective Cataract Surgery and Age-related Macular Degeneration Study

At-risk Right Eye of Participants

With No Known Cause of ERM

Eye-specific Cumulative

Incidence of ERM, % (95% CI)

At-risk Study Eye of Patients

After Cataract Surgery

Eye-specific Cumulative

Incidence of ERM, % (95% CI)

Standardized to the

BMES % (95% CI)

64-69 400 5.5 (3.5-8.2) 246 12.5 (8.9-17.6)

70-74 212 4.7 (2.3-8.5) 278 13.1 (9.5-18.0)

75-79 123 1.6 (0.2-5.8) 337 10.4 (7.5-14.5)

80þ 46 0.0 258 8.6 (5.6-13.2)

Total 781 4.4 (3.0-6.0) 1119 11.2 (9.4-13.4) 12.1 (8.6-16.9)

BMES ¼ Blue Mountains Eye Study; CI ¼ confidence interval; ERM ¼ epiretinal membranes.aExcluding participants with either substantial cataract or a secondary cause of ERM.

taken, and side-by-side grading of these images togetherwith adjudication for all questionable cases helped ascer-tainment of incident cases as well as progression and regres-sion of the cases. In addition, we used a similar gradingsystem and found similar results for post–cataract surgicalERM prevalence to that in the BMES and the BDES.1,4

These findings from photographic grading are applicableto clinical practice where OCT is not routinely performed.

In our study, patients who developed preretinal fibrosishad a similar proportion of visual impairment (VA <6/12)to those without preretinal fibrosis, suggesting that thepreretinal fibrosis cases detected in our surgical cohortwere likely at an early stage and had not yet had significantclinical consequence; we cannot exclude the possibility

VOL. 156, NO. 2 EPIRETINAL MEMBRANES IN EYES

that these lesions may progress to vision-impairing prereti-nal fibrosis in the future.15 Future studies to assess the long-term consequences on vision of persistent preretinal fibrosisare warranted.In summary, our findings show that 11% of surgical eyes

developed ERM, including 4% that developed preretinalfibrosis, over 3 years, after excluding the possibility ofdetection bias attributable to the presence of preoperativecataract. Of eyes with cellophane reflex detected shortlyafter surgery, 13.5% progressed to preretinal fibrosis overa 3-year period post phacoemulsification surgery. The inci-dence of ERM was higher in post–cataract surgical eyes ofthis surgical cohort compared to the idiopathic ERM inci-dence found in a general older population of Australians.

ALL AUTHORSHAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSUREOF POTENTIAL CONFLICTS OF INTEREST.C.S.F. received an international travel grant from the Association of Research in Vision and Ophthalmology (2011). P.M. serves on Advisory Boards forNovartis and Bayer, and has received consultancy fees and payments for lectures from these companies. J.J.W. is funded by a National Health & MedicalResearch Council Senior Research Fellowship (2005-2014). The rest of the authors declare no financial interests. The Australian Prospective CataractSurgery and Age-related Macular Degeneration Study was supported by the National Health & Medical Research Council, Australia (Grant No302010, 2004-2006), Retina Australia (2005) and NHMRC Centres for Clinical Research Excellence (Identification No 529923, 2009-2013). TheBMES is an Australian cohort study that was supported by the National Health & Medical Research Council, Australia (Grant No 932085, 974159,211069, and 457349). Both studies are registered with the Human Research Ethics Committees of the University of Sydney and the Sydney WestArea Health Service. Contributions of authors: study design (J.J.W., P.M.); conduct of the study (T.d.L., C.S.F.); data collection (C.S.F., T.d.L.,T.H.); management of the study (J.J.W., P.M.); statistical analyses (E.R.); analysis and interpretation of data (E.R., J.J.W., C.S.F.); drafting the manuscript(C.S.F., J.J.W.); review of the article (J.J.W., P.M., E.R., T.H., T.d.L., C.S.F.); and approval of the final version of the article, including changesmade in thesubsequent revision (J.J.W., P.M., E.R., T.H., T.d.L., C.S.F.). The authors wish to thankMsMireilleMoffitt for grading retinal photographs andMs KirstenJakobsen for editing the article and retinal photographs for publication.

REFERENCES

1. Mitchell P, SmithW, Chey T,Wang JJ, Chang A. Prevalenceand associations of epiretinal membranes. The Blue Moun-tains Eye Study, Australia. Ophthalmology 1997;104(6):1033–1040.

2. Fraser-Bell S, Ying-Lai M, Klein R, Varma R. Prevalence andassociations of epiretinal membranes in latinos: the LosAngeles Latino Eye Study. Invest Ophthalmol Vis Sci 2004;45(6):1732–1736.

3. McCarty DJ, Mukesh BN, Chikani V, et al. Prevalence andassociations of epiretinal membranes in the visual impairmentproject. Am J Ophthalmol 2005;140(2):288–294.

4. KleinR,KleinBE,WangQ,MossSE.Theepidemiologyof epire-tinal membranes. Trans Am Ophthalmol Soc 1994;92:403–425.

5. Appiah AP, Hirose T. Secondary causes of premacularfibrosis. Ophthalmology 1989;96(3):389–392.

6. Jahn CE, Minich V, Moldaschel S, et al. Epiretinalmembranes after extracapsular cataract surgery. J CataractRefract Surg 2001;27(5):753–760.

317AFTER CATARACT SURGERY

7. Cugati S, de Loryn T, Pham T, Arnold J, Mitchell P,Wang JJ.Australian prospective study of cataract surgery and age-related macular degeneration: rationale and methodology.Ophthalmic Epidemiol 2007;14(6):408–414.

8. Attebo K, Mitchell P, Smith W. Visual acuity and the causesof visual loss in Australia. The Blue Mountains Eye Study.Ophthalmology 1996;103(3):357–364.

9. Klein BE, Klein R, Linton KL, Magli YL, Neider MW. Assess-ment of cataracts from photographs in the Beaver Dam EyeStudy. Ophthalmology 1990;97(11):1428–1433.

10. Mitchell P, Cumming RG, Attebo K, Panchapakesan J. Prev-alence of cataract in Australia: the Blue Mountains eye study.Ophthalmology 1997;104(4):581–588.

11. Early Treatment Diabetic Study (ETDRS) Research Group.ETDRS Manual of Operations. Accession no. PB85–223006.Springfield,VA:NationalTechnical InformationService; 1985.

12. Asbell PA, Dualan I,Mindel J, Brocks D,AhmadM, Epstein S.Age-related cataract. Lancet 2005;365(9459):599–609.

13. Kawasaki R, Wang JJ, Mitchell P, Aung T, Saw SM,Wong TY. Racial difference in the prevalence of epiretinal

318 AMERICAN JOURNAL OF

membrane between Caucasians and Asians. Br J Ophthalmol2008;92(10):1320–1324.

14. Contreras I, Noval S, Tejedor J. Use of optical coherencetomography to measure prevalence of epiretinal membranesin patients referred for cataract surgery.Arch Soc Esp Oftalmol2008;83(2):89–94.

15. Fraser-Bell S, Guzowski M, Rochtchina E, Wang JJ,Mitchell P. Five-year cumulative incidence and progressionof epiretinal membranes: the Blue Mountains Eye Study.Ophthalmology 2003;110(1):34–40.

16. Mitchell P, Smith W, Attebo K, Wang JJ. Prevalence of age-related maculopathy in Australia. The Blue Mountains EyeStudy. Ophthalmology 1995;102(10):1450–1460.

17. Mitchell P, Smith W, Chang A. Prevalence and asso-ciations of retinal vein occlusion in Australia. The BlueMountains Eye Study. Arch Ophthalmol 1996;114(10):1243–1247.

18. Mitchell P, Smith W, Wang JJ, Attebo K. Prevalence of dia-betic retinopathy in an older community. The Blue Moun-tains Eye Study. Ophthalmology 1998;105(3):406–411.

AUGUST 2013OPHTHALMOLOGY

Biosketch

Calvin Sze-un Fong, MBBS, graduated in medicine from the University of New South Wales, Australia, in 2005. He

subsequently completed his medical internship at Westmead Hospital, in Sydney. He is currently undertaking a PhD at

the Centre for Vision Research, University of Sydney, under the supervision of Professor Jie Jin Wang and Professor

Paul Mitchell.

VOL. 156, NO. 2 318.e1EPIRETINAL MEMBRANES IN EYES AFTER CATARACT SURGERY