Objectives To analyse clinical features, systemic associations, treatment and visual outcomes of uveitis in children and adolescents at a tertiary centre in Tokyo.
Methods Clinical records of 64 patients under the age of 20 years who presented between 2001 and 2013 to the Ocular Inflammation Service of the Kyorin Eye Center, Tokyo were reviewed retrospectively.
Results Of the 64 patients, there was a predominance of girls (70%) and bilateral disease (81%). Mean age at presentation was 12.9 years (4–19 years). Mean follow-up was 46 months (3–144 months). Anterior uveitis was present in 56.3% of patients, panuveitis in 28.1% and posterior uveitis in 15.6%. No patients had intermediate uveitis. The most common diagnostic designation was unclassified uveitis (57.8%). Systemic associations were observed in 10.9% and no patients were diagnosed with juvenile idiopathic arthritis. Ocular complications were observed in 71.9% of patients, including optic disc hyperemia/oedema (40.6%), vitreous opacification (23.4%), posterior synechia (18.7%), increased intraocular pressure (17.1%) and cataract (14.1%). Six patients underwent intraocular surgery, five for cataract extraction and two for glaucoma control. Twelve patients (18.7%) received some form of systemic therapy either corticosteroids, immunosuppressive drugs or biologic agents. The percentage of eyes with a visual acuity of 1.0 or better was 87.1% at baseline, 91.3% at 6 months, 89.6% at 12 months and 87.5% at 36 months.
Conclusions The majority of children and adolescents who presented to us with uveitis had bilateral disease and no systemic disease associations. Only one-fifth of patients required systemic therapy to control their ocular inflammation, and most eyes had a good visual outcome.
- Child health (paediatrics)
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Children account for 5–10% of patients with uveitis seen at tertiary referral centres.1 ,2 It is believed that uveitis in children may be more severe than that of adults due to the development of vision-threatening ocular complications.1 ,3 The cause of uveitis in children can vary greatly depending upon the population being examined. For example, although anterior uveitis associated with juvenile idiopathic arthritis (JIA) is most common in the US and northern European countries, it is less frequent in Mediterranean and Middle Eastern countries.2 To date, there are few published reports in English on the manifestations of childhood uveitis from Asia, and none from Japan that we are aware of. In the present study, we analysed clinical features, systemic associations, treatment and visual outcomes of uveitis in children and adolescents in consecutive new patient referrals over a 14-year period at a tertiary centre in Tokyo.
A database was created for clinical record-keeping purposes of consecutive patients with intraocular inflammation (uveitis) referred to the Ocular Inflammation Service at the Kyorin Eye Center, Kyorin University Hospital in Tokyo, Japan. Our selection criteria for childhood and adolescent uveitis was referral before the patient age of 20 years between 2000 and 2013. Patients followed for less than 3 months were excluded. A total of 64 patients were identified for this study, with a mean age of 12.9 years (range 4–19 years).
The medical records of all patients were retrospectively reviewed. The data collected included age at onset of uveitis, gender, anatomic location of inflammation, specific diagnosis, visual acuity, ocular complications, ocular surgery, and use of local and/or systemic therapies. This study was conducted in accordance with the tenets of the Declaration of Helsinki and was approved by the Kyorin University Hospital Research Ethics Committee.
Using the anatomic location classification of the Standardised Uveitis Nomenclature (SUN) guidelines published in 2005, patients were categorised as having anterior uveitis, intermediate uveitis, posterior uveitis or panuveitis.4 Laboratory testing included blood and urine tests, chest X-ray, electrocardiogram and the tuberculin skin test (0.05 μg purified protein derivative in 0.1 mL, equivalent to 2.5 tuberculin units). Laboratory tests directed by uveitis findings (antinuclear antibodies, rheumatoid factor, angiotensin-converting enzyme level, and serologies for toxoplasma, toxocara, Bartonella henselae, herpes simplex virus and/or herpes zoster virus, and urine β2 microglobulin) were also performed when deemed necessary. Patients were referred to paediatric nephrology or rheumatology specialists for positive laboratory findings and/or systemic symptoms. The diagnosis of specific uveitic disorders or systemic disease associations was based on the combination of detailed clinical history, extensive review of systems, complete ophthalmologic examination, laboratory testing and the results of systemic consultations.
The SUN criteria were used to report duration and course of uveitis.4 Treatments specific for uveitis were recorded, including corticosteroids (local and systemic), non-steroidal anti-inflammatory drugs, antibiotics, methotrexate (MTX), cyclosporine, colchicine and infliximab. The frequency of topical corticosteroid (betamethasone) eyedrops before and after initiation of MTX treatment was analysed to evaluate the effectiveness of systemic MTX in patients with chronic uveitis. Complications were recorded if they occurred at any time point, including band keratopathy, posterior synechia, iris atrophy, cataract, ocular hypertension (defined as intraocular pressure >21 mm Hg), vitreous opacification, maculopathy, macular oedema, epiretinal membrane, chorioretinal atrophy, optic disc hyperemia/oedema and central retinal vein occlusion. The number of patients who developed ocular complications was calculated at baseline, 6 months, and at 1, 2 and 3 years. Ocular surgery was divided by indication. For statistical analysis, nominal data were analysed using the Fisher exact probability test. The frequency of topical corticosteroid eyedrops before and after initiation of MTX treatment was analysed using the Mann–Whitney U test. A p value <0.05 was considered to be statistically significant.
The distribution of uveitis by age and gender is shown for all 64 patients in figure 1. Of the 64 patients, 45 (70.3%) were girls and 19 (29.7%) were boys. The mean age at onset was 12.9±4.4 years for all patients (range 4–19 years); 11.0 years for the boys (5–18 years) and 13.7 years for the girls (4–19 years). All patients were ethnically Japanese. The mean follow-up was 46 months (range 3–144 months).
As shown in figure 2, primary presenting symptoms included red eye (51.8%), reduced vision (28.6%), hazy vision (19.6%), ocular pain (17.9%), photophobia (8.9%), itching (8.9%) and tearing (3.6%). Over one-half of patients presented with anterior uveitis and no patients had intermediate uveitis (table 1). Among children 0–15 years of age (childhood group), 68.3% had anterior uveitis, while this proportion was 34.8% among adolescents 16–19 years of age (adolescent group). Conversely, the proportion of posterior and panuveitis in the adolescent group was higher compared with that in the childhood group (table 1). Overall, 81.2% of patients had bilateral disease and 19.8% had unilateral disease.
As shown in table 2, the three most common diagnostic classifications were unclassified uveitis (57.8%), juvenile chronic iridocyclitis without evidence of JIA (henceforth referred to as juvenile chronic iridocyclitis, 17.2%), and neuroretinitis (6.2%). Among the 11 patients with juvenile chronic iridocyclitis, 4 patients (36.4%) were positive for antinuclear antibody. The percentage of patients with juvenile chronic iridocyclitis was 22.0% in the childhood group and 8.7% in the adolescent group. Systemic associations were observed in 10.9% of patients, the most common being Behçet's disease in two patients (3.1%) and tubulointerstitial nephritis and uveitis (TINU) in two patients (3.1%). No patients were diagnosed with JIA.
Of patients in whom disease activity could be classified, the duration was most commonly persistent (39 patients, 60.9%) as opposed to limited (25 patients, 39.1%). A chronic disease course (34 patients, 53.1%) was more common than an acute (25 patients, 39.1%) or recurrent (5 patients, 7.8%) disease course.
Inflammation was successfully controlled in 38 patients (59.4%) using topical corticosteroids alone. No patients received local injection therapy. Twelve patients (18.7%) received some form of systemic therapy: corticosteroids, immunosuppressive drugs and/or infliximab (table 3). Six of these patients had juvenile chronic iridocyclitis and both patients who received infliximab were diagnosed with Behçet's disease. MTX was the most commonly prescribed systemic drug, administered in eight patients (12.5%) (six patients with juvenile chronic iridocyclitis, one patient with TINU and one patient with unclassified uveitis). Follow-up data were available in six patients (12 eyes) of the eight patients treated with MTX.
The median initial dose of MTX was 4 mg weekly (range 4–6 mg) and the median maximum dose was 9 mg weekly (range 4–16 mg). MTX was administered orally in all cases and in an attempt to reduce topical corticosteroid use when steroid-induced ocular hypertension was suspected. The median duration of MTX therapy was 36 months (range 14–94 months). The topical corticosteroid (0.1% betamethasone) frequency significantly decreased from a mean of 6.0 times a day before initiation of MTX therapy to 3.2 times a day at 6 months (p=0.0148) and 1.4 times a day at 12 months (p=0.0325) after initiation of MTX therapy (figure 3). All patients were able to continue on MTX therapy without adverse effects. Cyclosporine was used in three patients each (Behçet's disease in two patients, idiopathic panuveitis in one patient).
Ocular complications were observed in 71.9% of patients either at presentation or during the follow-up period. As shown in table 4, the most common complications were optic disc hyperemia/oedema (40.6%), vitreous opacities (23.4%), posterior synechia (18.8%), increased intraocular pressure (17.1%), cataract (14.1%) and chorioretinal atrophy (12.5%). Macular complications such as macular oedema and epiretinal membrane were uncommon. Band keratopathy was observed in only two patients (3.1%).
Six patients underwent intraocular surgery, five eyes for cataract extraction (three eyes with juvenile chronic iridocyclitis, one eye with TINU, one eye with Behçet's disease), and two eyes for glaucoma control (one eye each with juvenile chronic iridocyclitis and unclassified uveitis). A recent report suggested that a major risk factor for development of early cataract requiring surgery in children with JIA-associated uveitis is the presence of posterior synechia at the time of diagnosis of uveitis.5 We performed an analysis to determine whether the presence of posterior synechia at the time of diagnosis of uveitis was associated with development of cataract requiring surgery in our patients. Three of 17 eyes (17.6%) with posterior synechia at the initial visit developed cataract requiring surgery, while the same was true for only 2 of the 97 eyes (2.1%) without synechia at initial visit. This difference was statistically significant (p=0.022).
Figure 4 shows the best-corrected visual acuity (BCVA) as measured in decimal units at specific follow-up intervals for all patients in whom these data were available. The percentage of eyes with a BCVA of 1.0 or better was 87.1% at baseline, 91.3% at 6 months, 89.6% at 12 months, 87.5% at 36 months, 80.0% at 60 months and 93.1% at final visit.
Paediatric uveitis represents 5–10% of the general uveitis population.1 ,2 ,6 Previous series from North America and Europe have reported that anterior uveitis accounts for 35–50%, posterior uveitis 15–25%, intermediate uveitis 10–20% and panuveitis 10–20% of all paediatric patients with uveitis at tertiary centres.2 In the present study, there was a predominance of anterior uveitis (56.2%) followed by panuveitis (28.1%) and posterior uveitis (15.6%), whereas no patients were diagnosed with intermediate uveitis. The most common systemic disease was Behçet's disease.
A large series by Goda et al7 (published in the Japanese language) of 104 Japanese children before the age of 16 with uveitis consisting entirely of referral patients reported a low frequency of intermediate uveitis (4.8%). In contrast, intermediate uveitis accounts for 4–15% of uveitis at referral practices in North America and predominantly affects children.8 A significant association between intermediate uveitis and HLA-DR15 has been reported, and this HLA type is less common in Japan compared with that in predominantly Caucasian countries.9–11 In addition, multiple sclerosis (MS) is known to be associated with intermediate uveitis, and the prevalence of MS is lower in Japan compared with North America and Europe.9 ,12–14 Therefore, differences in genetic background and/or associated diseases may be a major cause of the absence of intermediate uveitis in our study. Interestingly, surveys from Turkey and India have shown that intermediate uveitis accounts for only 8–11% of childhood patients, and this may represent a rate in between Japan and North America/Europe.15 ,16
In the Goda study of Japanese children, the frequency of anterior uveitis was 48.1%, posterior uveitis 4.8% and panuveitis 42.3%.7 In comparison, we observed a higher frequency of posterior uveitis and a lower frequency of panuveitis. This may be due to differences in criteria used for categorising uveitis, as well as the fact that we included patients up to the age of 19 whereas the Goda study only included patients up to the age of 15. As shown in tables 1 and 2, the ratio of posterior uveitis in the adolescent group was higher compared with that in the childhood group, and thus this may account for the greater numbers of patients with posterior uveitis in our study. With regards to panuveitis, it is important to note that the most common specific diagnosis in the Goda report was sarcoidosis, and this may have contributed to a high number of patients with panuveitis in that study. Although we had no patients diagnosed with sarcoidosis, both our previous research and nationwide surveys have demonstrated that the most common specific disease classification for uveitis in patients of all ages is sarcoidosis.6 ,17 ,18 Therefore, it is possible that sarcoidosis was underdiagnosed in our survey.
No specific aetiology or disease association could be determined in the majority of patients in our study. We acknowledge that this may be due, in part, to the absence of specific testing in some cases, such as for herpes viruses by PCR or Goldmann–Witmer coefficient analysis of aqueous humour samples. However, paracentesis is much more difficult to perform in children than in adults, and we did not believe that the severity of disease justified this invasive procedure in our young patients. Furthermore, as already stated, sarcoidosis may have been underdiagnosed, particularly when ancillary testing such as chest CT and gallium scan were not pursued in these children.
Notably, the results of our survey differs greatly with reports from North America and Europe where JIA was the most commonly identified specific diagnosis, accounting for 20–40% of uveitis in children younger than 16 years of age.19–23 However, JIA was found to be a less frequent cause of paediatric uveitis in Mediterranean and Middle Eastern countries, as well as in India.16 ,24–26 In our study, there was not a single patient with uveitis associated with JIA. Of note, the Goda study of Japanese children included only one patient with JIA-associated uveitis.7 Several smaller series from Japan (published in Japanese) have reported that JIA was present in only 0.9–10.6% of patients with paediatric uveitis at referral centres.27–29 Part of the reason for the small numbers of JIA-associated uveitis in Japan must certainly be due to the low prevalence of JIA itself in Japan, reported to be 9.7 per 100 000, compared with 7–150 per 100 000 in countries with populations of predominantly European descent.30–32 Furthermore, the oligoarthritis type of JIA, associated with a higher risk of iridocyclitis,33 is much less common in Japan compared with in North America and Europe where it is the most common type of JIA.32 ,34 Together, these may explain the relative rarity of children with JIA-associated uveitis in Japan.
Our study demonstrated that over 80% of patients maintained a visual acuity of 1.0 or better during the follow-up period. In one series at another Japanese institution, 73% of eyes with paediatric uveitis had a visual acuity of 0.5 or better at the last visit.27 Smith and colleagues reported that the proportion of children with moderate visual compromise of visual acuity less than or equal to 20/50 was 26%, and that study suggested that posterior infectious uveitis was a risk factor for poor visual outcome.23 It has been reported that paediatric posterior infectious uveitis with macular lesions, such as due to toxoplasmic retinochoroiditis, has a poor visual prognosis.35 In the present study, although the prevalence of posterior uveitis was 15.6%, the frequency of maculopathy (4.7%) and macular oedema (3.1%) was low. In addition, several reports have demonstrated that patients with paediatric uveitis and JIA have a higher risk of specific, common uveitis complications compared with patients with other etiologic diagnoses.2 ,36 ,37 Thus it may be that a lower proportion of macular complications, combined with the absence of JIA-associated uveitis, was related to the relatively good visual outcomes we observed.
In terms of treatment, while 60% of patients received topical corticosteroids alone, 12 patients (18.7%) received some form of systemic therapy: corticosteroids, immunosuppressive drugs or biologic agents. In our hands, systemic treatment was often initiated with the goal of reducing topical corticosteroid use when steroid-induced glaucoma was suspected. Conventional immunosuppressive drugs were given in most cases, with MTX being the most common drug administered in eight patients (12.5%), six patients with juvenile chronic iridocyclitis, one patient with TINU and one patient with idiopathic panuveitis. Several reports have demonstrated that low-dose MTX is effective for paediatric chronic uveitis and we found low-dose MTX to be useful in patients with scleritis.38–40 Consistent with this, we found that the frequency of topical corticosteroids could be significantly reduced after initiation of MTX treatment. Importantly, none of our patients required discontinuation of MTX due to side effects. Infliximab was used in two adolescent patients, both of whom had Behçet's disease, the only ocular disease for which infliximab use has been approved by the national healthcare system in Japan. Infliximab therapy was well tolerated in these patients and no side effects were observed.
It has been reported that a risk factor for the early development of cataract requiring surgery in children with JIA-associated uveitis is the presence of posterior synechia at the time of diagnosis of uveitis.5 We also found that the presence of posterior synechia at the time of uveitis diagnosis was statistically correlated with development of cataract requiring surgery. Moreover, three of the five eyes (60%) that underwent cataract surgery had juvenile chronic iridocyclitis, supporting a relation between chronic anterior segment inflammation, posterior synechia formation and cataract formation, regardless of etiologic diagnosis.
Possible biases and limitations of this study were that the data were examined retrospectively and collected from a single institution. Furthermore, our study was conducted in a subspecialty clinic at a university hospital and therefore does not reflect the entire spectrum of paediatric uveitis observed and treated in the community. Moreover, the choice of treatment modalities was greatly influenced by health insurance policies in Japan and personal preferences of the ocular inflammation specialists involved. In addition, comparisons to other surveys of paediatric uveitis are complicated by the inclusion of different age ranges of children and the use of different diagnostic definitions for disease. Regardless, we believe that our results add new and important knowledge regarding the clinical features of Japanese children and adolescents with uveitis.
In conclusion, the majority of children and adolescents who presented to us with uveitis had bilateral disease and no systemic disease associations. One-fifth of patients required systemic therapy and most eyes had a good visual outcome. A prospective multicentre study would aid in further delineating the disease types and clinical features of paediatric uveitis in Japan.
Competing interests None declared.
Ethics approval Research Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.