Journal 3
Journal 3
PURPOSE: To determine the risk factors for the inflammation recurrences and ocular complications. We
development of cataract in children with uveitis of any suggest that controlling the inflammation, even using
etiology. higher doses of systemic and topical corticosteroids, is
DESIGN: Cohort study. of importance in preventing ocular complications, such
METHODS: Two hundred forty-seven eyes of 140 chil- as cataract. (Am J Ophthalmol 2017;177:139143.
dren with uveitis were evaluated for the development of 2017 Elsevier Inc. All rights reserved.)
vision-affecting cataract. Demographic, clinical, and
treatment data were collected between the time of presen-
U
tation and the first instance cataract was recorded or find- VEITIS ACCOUNTS FOR 1015% OF BLINDNESS IN
ings at final follow-up. Main outcome measures included the developed world.1 Although pediatric uveitis
the prevalence of cataract and distribution by type of uve- is relatively uncommon, accounting for only
itis, incidence of new onset cataract time to cataract 510% of all uveitis cases,2 it affects young patients, who
development, and risk factors for the development of in most cases are otherwise healthy. Vision loss results
cataract. from ongoing inflammation that leads to ocular structural
RESULTS: The prevalence of cataract in our cohort was changes, such as cataract, corneal opacities, optic neurop-
44.2% and was highest among eyes with panuveitis athy, and retinal lesions. The most common causes of
(77.1%), chronic anterior uveitis (48.3%), and interme- vision loss in children with uveitis are cataract, glaucoma,
diate uveitis (48.0%). The overall incidence of newly and chronic cystoid macular edema (CME).2,3 In addition,
diagnosed cataract was 0.09 per eye-year, with an esti- any chronic visual obstruction can result in the
mated 69% to develop uveitis-related cataract with development of amblyopia in younger children, with
time. The main factors related with cataract development vision loss persisting after the inciting cause is treated.4
were the number of uveitis flares per year (hazard ratio Such changes, together with the need for long-term
[HR] [ 3.06 [95% confidence interval {CI}, 2.15 treatment and continuous monitoring, can have a profound
4.35], P < .001), cystoid macular edema (HR [ 2.87 impact on their development, independence, and
[95% CI, 1.415.82], P [ .004), posterior synechia at education.
presentation (HR [ 2.85 [95% CI, 1.535.30], The prevalence of cataract in eyes with uveitis ranges
P [ .001), and use of local injections of corticosteroids from 2064%,47 and it is the most common
(HR [ 2.37 [95% CI, 1.184.75], P [ .02). Treat- complication of uveitis in children,8 occurring in approxi-
ments with systemic and topical corticosteroids were mately 35% of children with juvenile idiopathic arthritis
not significant risk factors. (JIA)-associated uveitis9 and increasing < _80% in
CONCLUSIONS: In this study, we found that develop- adults.10,11 Cataract progression can be the result of
ment of cataract is common among pediatric eyes with persistent intraocular inflammation,12,13 can be caused by
uveitis and is most strongly related to the extent of surgery for uveitis complications (eg, trabeculectomies
and repair of retinal detachments), or can be a
consequence of uveitis treatment, particularly the use of
local or systemic corticosteroids.1417 It results in reduced
Supplemental Material available at AJO.com.
Accepted for publication Feb 22, 2017. visual acuity and can have a detrimental effect on the
From Moorfields Eye Hospital (T.B.-H., L.S., S.L., O.T.-N.), London, development and academic achievements of these
United Kingdom; Department of Ophthalmology (T.B.-H., G.H.), children.18
Galilee Medical Center, Naharia, Israel; UCL Institute of
Ophthalmology (S.S.-G., L.S., S.L., O.T.-N.), London, United Studies have examined risk factors for the development
Kingdom; Department of Ophthalmology, Rabin Medical Center (M.K., of cataract among children with JIA-associated uveitis,
Y.S.), Petah-Tikva, Israel; Sackler School of Medicine (M.K., R.F.), Tel identifying risk factors such as the presence of posterior
Aviv University, Tel Aviv, Israel; Pediatric Ophthalmology Unit (R.F.),
Schneider Childrens Medical Center of Israel, Petah-Tikva, Israel; and synechiae (PS) at presentation,12,19 the use of systemic
the Department of Ophthalmology (O.T.-N.), Bnai Zion Medical corticosteroids,13 topical corticosteroid therapy exceeding
Center, Technion, Israel Institute of Technology, Haifa, Israel. 3 drops a day,12 or persistent, uncontrolled active inflam-
Inquiries to Oren Tomkins-Netzer, Institute of Ophthalmology,
University College of London, 162-165 City Rd, London EC1V 2PD; mation,3 while early treatment with methotrexate delayed
e-mail: o.tomkins-netzer@ucl.ac.uk cataract progression.19 However, JIA is a unique cause of
METHODS
THIS IS A RETROSPECTIVE STUDY CONDUCTED AT MOOR-
fields Eye Hospital, London, United Kingdom (ethical
approval for data collection ROAD16039, visual loss in FIGURE. Cumulative incidence analysis for cataract develop-
uveitis), and at the Schneider Childrens Medical Center ment among phakic eyes with no cataract at presentation.
of Israel/Rabin Medical Center, Petah Tikva, Israel Median time to cataract development was 96 months (95% con-
(ethical approval 0307-14-RMC). Data were collected fidence interval, 57.0135.0), with an overall incidence of
from 20002014. The study adhered to the Declaration newly diagnosed cataract of 0.09 per eye-year (95% confidence
of Helsinki and all state laws. The pediatric uveitis clinics interval, 0.070.11).
were tertiary referral centers with children referred by
ophthalmologists or pediatric rheumatologists, and in addi-
tion presented directly to the accident and emergency as presence of any intraocular cells or flare. Flare-ups
departments. were treated using local or systemic immunosuppression
Patients were included if they were <18 years of age at treated to achieve rapid disease control.
the time of uveitis diagnosis. Eyes were excluded if they The main outcome measures were the prevalence and
had other conditions that could cause cataract (eg, trauma incidence of cataract for the entire cohort and the time
or congenital cataract). For estimating the time and risk to and risk factors for cataract development among those
factors for cataract development, we excluded any patients with no cataract at diagnosis.
who had cataract extraction surgery before the diagnosis of
uveitis, had documented cataract at presentation, or were STATISTICAL ANALYSIS: Analyses were performed on a
followed for <6 months. Information about the patients per-eye basis. The time to cataract development was calcu-
was gathered until the time they developed cataract or, if lated using KaplanMeier survival analysis. A multivariate
none developed, until their last follow-up visit. Cox regression analysis was conducted to assess the hazard
Patients clinical details and treatment information was ratio (HR) and 95% confidence intervals (CIs) for risk fac-
gathered from their clinical notes, during the time interval tors to developing cataract, for which we used all variables
from presentation to the diagnosis of cataract. The that were significant (P < .05) in a univariate analysis. All
following information was gathered: sex, age at presenta- Cox regression models used robust variance estimation to
tion, anatomic diagnosis, etiology when identified after account for correlation between eyes in patients with bilat-
relevant investigations, presence of PS at presentation, eral uveitis. Analyses were performed using SPSS statistical
CME at any time, use of systemic corticosteroids, second- software (version 21; IBM, Chicago, IL). Results are
line immunosuppressive therapy, use of local corticosteroid presented as averages 6 standard error of mean. P < .05
injections (eg, orbital floor injections, intravitreal triam- was considered significant.
cinolone acetate injections, and dexamethasone implants),
the number of weeks that the patient was treated with
topical corticosteroids of >3 drops per day, and the number
of uveitis flare-ups. RESULTS
The type of uveitis was classified using the criteria of the
Standardization of Uveitis Nomenclature working group.20 OUR STUDY INCLUDED 247 EYES OF 140 PEDIATRIC PATIENTS
Cataract was defined as any opacity of the lens related to a (59% female) diagnosed with uveitis. The average age at
nontransient decrease in vision. Uveitis flare-ups were presentation was 10.3 6 0.4 years. There were 107
defined as any event of increased intraocular inflammation (76.4%) patients with bilateral uveitis. The mean follow-
that required treatment. Active inflammation was defined up time was 51.6 6 3.4 months (range, 6261 months),
and the prevalence of cataract in the entire cohort was cinolone acetate, and 9 months for dexamethasone
44.2% of eyes and occurred in 12.9% of eyes with acute implants. Local treatment was repeated as clinically
anterior uveitis, 48.3% of those with chronic anterior uve- required, when inflammation was active.
itis, 48.0% with intermediate uveitis (IU), 16.7% with pos- We examined factors related with cataract development
terior uveitis (PostU), and 77.1% with panuveitis (PanU). in these patients (Table). Among types of uveitis, PanU
The most common etiologies related to cataract were idio- was found to be a statistically significant risk factor for cata-
pathic (48.6%), JIA (12.9%), antinuclear antibodyposi- ract development (P .02), while CAU was related to an
tive vasculitis with no evidence of joint involvement increased risk of cataract development but did not reach
(8.6%), infections (eg, acute retinal necrosis, tuberculosis, statistical significance (P .07). We examined separately
toxoplasmosis, and HIV infections; 7.8%), idiopathic pos- clinical properties that might be related to cataract devel-
terior uveitis (multifocal choroiditis and acute multifocal opment and found that in the univariate analysis all clin-
placoid pigment epitheliopathy; 5.7%), human leukocyte ical factors, apart from sex, laterality, and treatment with
antigen-B27positive (5%), and sarcoidosis (4.3%). immunosuppressive therapy, were found to be significantly
At presentation, 164 eyes of 94 patients (62.8% female) related to an increased risk of developing cataract. Howev-
were phakic with no cataract. We examined this cohort to er, once all factors were adjusted for, only the number of
determine risk factors for the development of cataract after flare-ups per year (HR 3.06 [95% CI, 2.154.35]; P <
the diagnosis of uveitis. The mean age at presentation was .001), PS (HR 2.85 [95% CI, 1.535.30]; P .001), local
11.2 6 0.4 years (range, 318 years), and 70 children corticosteroid injections (HR 2.37 [95% CI, 1.184.75];
(74.5%) had bilateral uveitis. During follow-up, 61 eyes P .02), and development of a CME remained significant
developed cataract (37.2%). (HR 2.87 [95% CI, 1.415.82]; P .004; Table). Use of
The median time to develop cataract was 96 months systemic corticosteroids and the duration of time of >3
(95% CI, 56.9135.1; Figure), with an overall incidence drops of topical corticosteroids per day lost significance.
of 0.1 cases per eye-year (95% CI, 0.070.1), and we esti-
mated that < _69% of eyes would ultimately develop cataract
by 129.5 months (95% CI, 104.1154.9). The distribution
of newly diagnosed cataract by uveitis type was 28 eyes DISCUSSION
(17.1%) acute anterior uveitis, 52 eyes (31.7%) CAU, 23
eyes (14%) PanU, 50 eyes (30.5%) IU, and 11 eyes WE EVALUATED RISK FACTORS FOR THE DEVELOPMENT OF
(6.7%) PostU. Forty-eight eyes (29.3%) had PS at presen- cataract in children with uveitis. Complications of uveitis
tation, and 43 eyes (26.2%) developed CME during follow- are considered more prevalent in children than in adults
up. Forty-eight patients (87 eyes, 53%) were treated with because of delayed diagnosis and the difficulty in clinical
systemic corticosteroids for a period of > _6 months, 37 examination and treatment, with the most common
patients (67 eyes, 40.9%) received immunosuppressive complication being cataract.5,21
therapy, 24 eyes (14.6%) received orbital floor injections The prevalence of cataract based on the published liter-
of corticosteroids, 15 eyes (9.1%) received intravitreal ature ranges from 2064% of eyes.5,6,8 Studies focusing on
triamcinolone acetate, and 7 eyes (4.3%) were given dexa- JIA-related uveitis found a lower prevalence than studies
methasone implants. Injections lasted on average 6 weeks examining children with diverse causes of uveitis. Our
for orbital floor injections, 12 weeks for intravitreal triam- results correspond with the higher values of this range
FUNDING/SUPPORT: NO FUNDING OR GRANT SUPPORT. FINANCIAL DISCLOSURES: SUE LIGHTMAN HAS RECEIVED CONSUL-
tancy fees from Allergan, GSK, 4Sight Bayer, AbbVie, and Paraxcel, and has been on the advisory boards of Allergan, Bayer, AbbVie, Santen, and
GSK. Michal Kramer has been on the advisory board of AbbVie. Oren Tomkins-Netzer has received consultancy fees and been on the advisory board
of AbbVie. The following authors have no financial disclosures: Tamar Blum-Hareuveni, Sophie Seguin-Greenstein, Guy Hareuveni, Yael Sharon, Ronit
Friling, and Lazha Sharief. All authors attest that they meet the current ICMJE criteria for authorship.