Background
Childhood ocular rosacea is a rare but often misdiagnosed disorder associated with chronic blepharitis and, in many cases, phlyctenular keratoconjunctivitis
[
1]. It is alternatively named blepharokeratoconjunctivitis
[
2,
3] or staphylococcal phlyctenular keratoconjunctivitis. Lid involvement is characterized by chronic posterior and anterior blepharitis with frequent chalazia. Ocular inflammation is often unilateral with chronic conjunctival hyperemia; phlyctenules on the conjunctiva, the limbus, or the cornea; and inferior punctate keratopathy. Occasionally, corneal involvement is more severe with predominantly inferior (sub) epithelial infiltrates, stromal ulcers, neovascularization, and scars, leading to visual sequelae in 17% of cases
[
4]. Skin involvement is inconstant.
The primary mechanism of childhood ocular rosacea is chronic meibomitis. Staphylococcal infection in the meibomian glands is probably secondary and may induce conjunctival and corneal inflammation by a T lymphocyte immune response against bacterial parietal antigens or toxins. The phlyctenules and corneal infiltrates are the result of a type IV cell-mediated late hypersensitivity reaction according to the Gell and Coombs' classification
[
5].
The treatment of childhood ocular rosacea is based on its presumed physiopathogenic mechanisms and involves improving meibomian gland function, controlling bacterial proliferation with lid hygiene and antibiotics, and inhibiting T cell-mediated inflammation with topical steroids or cyclosporine. Daily lid hygiene is the cornerstone of ocular rosacea treatment, with oral antibiotics required in case of lid hygiene failure or in severe forms of the disease. Second-generation tetracyclines are used in children over 8 years old, and erythromycin, azithromycin, or metronidazole are used in younger patients
[
6‐
8]. Topical antibiotics are often used, but very few publications support their efficacy
[
3]. Topical steroids are very effective but should only be used in case of acute corneal inflammation due to the risk of iatrogenic complications. Cyclosporine 0.5% to 2% eye drops is an interesting alternative in children with steroid-dependent disease
[
9].
Azithromycin is a macrolide antibiotic characterized by a broad antibacterial spectrum, a long half-life related to its tissue and cell penetration
[
10], and anti-inflammatory properties
[
11]. Oral azithromycin is an effective treatment in ocular rosacea
[
12] and also in cutaneous involvement of rosacea (with efficacy comparable to doxycycline)
[
13].
In this retrospective study, we assessed the efficacy of topical 1.5% azithromycin eye drops in children with ocular rosacea and phlyctenular blepharokeratoconjunctivitis.
Discussion
Childhood ocular rosacea is believed to be the consequence of chronic meibomian gland dysfunction. Secondary staphylococcal infection of the lid margin is often present and is probably responsible for the corneoconjunctival inflammation by means of an immunological specific T cell response. This combined mechanism explains the efficacy of both topical anti-inflammatory drugs (steroids or cyclosporine) and systemic antibiotics such as macrolides or tetracyclines.
Azithromycin is a new treatment option for cutaneous rosacea and seems to be as effective as tetracyclines
[
13], and its efficacy in treating ocular rosacea has been demonstrated when administered orally
[
12].
Azithromycin eye drops are now available for the treatment of bacterial conjunctivitis, and in this indication, azithromycin 1.5% eye drops given twice daily for 3 days is as effective as tobramycin 0.3% eye drops four times a day for 7 days
[
14].
Our study showed that topical azithromycin 1.5% is a very effective treatment of ocular rosacea in children. Its efficacy on conjunctival and corneal inflammation is remarkable. However, because of the delayed action on corneal inflammation, prior clinical experience has shown that very severe cases with vision threat should be concomitantly treated with topical steroids and/or cyclosporine. Oral erythromycin is usually prescribed in childhood ocular rosacea
[
15]. In this study, in one patient who had been unsuccessfully treated with oral erythromycin, replacement by azithromycin resulted in a complete control of inflammation. Our previous treatment strategy was to use oral erythromycin as a first-line treatment in combination with lid hygiene. However, our early experience with topical azithromycin showed us that this local treatment was superior to systemic erythromycin. This is why we decided to use, since then, topical azithromycin as a first-line therapy. Topical azithromycin has several advantages over oral antibiotics. Azithromycin has a very long half-life, with significant tissular accumulation and concentrations. In a rabbit model, after administration of 1% azithromycin ophthalmic solution, significant concentrations are detected in the tears, conjunctiva, cornea, and lids for as long as 6 days in the lids
[
10]. Concentrations in the tears, conjunctiva, and cornea remained above minimal inhibitory concentrations for respectively 7, 17, and more than 24 days after instillation of Azyter® (Thea Laboratories, Clermont-Ferrand, France) twice a day (bid) for 3 days
[
16]. This is why we chose a discontinuous treatment regimen which was very simple for the patients; one treatment corresponds to one box of eye drops, and so the initial scheme of three treatments per month tapered to two and then one per month was convenient, easy to understand, and well accepted by the children and their parents. The initial dosing (3 days three times monthly) was chosen because of the 7-day persistence of azithromycin in the tears. The tapering scheme was chosen empirically. Oral erythromycin has to be taken twice daily for a continuous period of several weeks. In the 16 patients studied, topical azithromycin did not induce systemic side effects, whereas oral erythromycin frequently induces gastrointestinal troubles. Oral metronidazole has also been proposed as a treatment for ocular rosacea, but while it seems to be an effective treatment, the fact that short-course therapy is required in order to avoid peripheral neuropathy may result in more frequent relapses
[
6].
Few publications reported the efficacy of topical antibiotics in this disease. In a retrospective study by Viswalingam et al., topical chloramphenicol eyedrops four times daily for 1 month and chloramphenicol ointment at night for 4 months were successfully prescribed in mild to moderate cases
[
3]. Azithromycin ophthalmic solution 1% has been evaluated in posterior blepharitis in adults in two small studies
[
17,
18]. In one study, meibum quality and lid margin redness improved more in patients treated with lid hygiene and azithromycin for 2 weeks than in controls treated with only lid hygiene. In the second study, symptoms, meibum quality, and lid margin redness improved after treatment with lid hygiene and azithromycin for 4 weeks. While we did not specifically analyze the posterior lid margin in our study, the effect of topical azithromycin on anterior blepharitis was noticeable, and no chalazia were noted.
Topical steroids are frequently prescribed in children with ocular rosacea and phlyctenular blepharokeratoconjunctivitis, but relapses occur after the end of the treatment in up to 40% of cases
[
15]. Treatment duration is a key element for avoiding recurrences. In our study, treatment was stopped after 4 to 10 months, and no recurrence was observed during a median follow-up of 11 months (8 to 14 months). Long-term treatments are usually required to control the inflammation. A complete disappearance of the phlyctenules and corneal infiltrates is necessary before stopping the treatment, which was attained in 3 to 6 months in the majority of children included in this study with topical azithromycin. This minimal treatment duration is comparable with topical cyclosporine
[
9] or topical chloramphenicol
[
3]. Such duration is not acceptable with topical steroids because of the risk of ocular complications. In our series, topical steroids were withdrawn in all patients when topical azithromycin was prescribed, and none of the patients required additional steroid therapy thereafter.
In a previous study, we showed that topical 2% cyclosporine is a very potent treatment of phlyctenular blepharokeratoconjunctivitis
[
9]. The efficacy of topical azithromycin was sufficient in most cases, but as clinical experience indicates that topical cyclosporine may have greater efficacy, patients with very severe corneal inflammation were prescribed topical cyclosporine as first-line treatment.
A biphasic effect was observed in this study. The effect on ocular redness was very fast, within a month, whereas phlyctenules and corneal infiltrates took several months to heal. This may point to different mechanisms. Ocular redness might be related to bacterial toxin production. Phlyctenules and corneal infiltrates are thought to be caused by a type IV cell-mediated delayed hypersensitivity reaction against parietal staphylococcal antigens
[
5]. In addition to its antibiotic effect, azithromycin also has anti-inflammatory effects. Indeed, it has been shown to reduce the production of the pro-inflammatory cytokines IL-12 and IL-6 by macrophages
in vitro
[
11]; to suppress matrix metalloproteases 2 and 9, nuclear transcription factor NFkB, and Toll-like receptor 2 in corneal epithelial cells
in vitro
[
19,
20]; and to inhibit macrophage and dendritic cell migration after corneal burn in a mouse model
[
21].
The safety profile of topical 1.5% azithromycin seems good in children suffering from ocular rosacea. Discontinuous treatment with topical 1.5% azithromycin was well tolerated. Local intolerance was seen in a small number of cases and was usually mild, except in two patients who stopped the treatment after 5 and 10 months, respectively. Comparable side effects have been reported in the literature
[
17,
18].
The main limitation of our study is the small number of patients, but rosacea in children is infrequent, and the largest series published to date included 44 children
[
3]. The retrospective nature of this study is also an important limitation. A multicenter prospective controlled study should be the next step for evaluating azithromycin eyedrops in childhood ocular rosacea.
Methods
Design
This is a retrospective, non-comparative, interventional case series.
Methods
We reviewed the files of all 16 consecutive children with rosacea and phlyctenular blepharokeratoconjunctivitis who were treated with topical 1.5% azithromycin eye drops between January 2009 and March 2010 in the Department of Ophthalmology at Bichat Hospital and Fondation A. de Rothschild, Paris, France. Written informed consent was obtained from the patients and from their parents for publication of this report. This study was approved by the Ethics Committee of Foundation A de Rothschild.
Treatment strategy
Patients presenting with ocular rosacea were initially treated with lid hygiene once daily, and frequent normal saline instillations were initially prescribed alone for at least 1 month. Patients with sight-threatening corneal infiltrates approaching the visual axis were treated with steroids and cyclosporine 2% eye drops and were not treated with azithromycin; therefore, they were not included in the study.
Azithromycin 1.5% eye drops (Azyter®, Thea Laboratories) was administered only to patients who did not respond to lid hygiene. A three-day treatment (1 drop bid) was initially prescribed every 10 days (i.e., three treatments monthly) for at least 1 month. According to clinical efficacy, treatment was tapered to one treatment every 15 days and then one treatment monthly.
Follow-up
Patients were examined after 1 month and every 1 to 2 months thereafter.
The main outcome measures were ocular redness (reported by the patients), occurrence of chalazia (reported by the patients), conjunctival bulbar hyperemia, conjunctival phlyctenules, corneal inflammation (phlyctenules and infiltrates) and epitheliopathy, and lid margin inflammation (blepharitis). All measures were graded on scales of 0 to 4. Ocular rosacea was considered to be controlled if all items except lid margin inflammation had a score of 0. Ocular and systemic adverse events were also reported.
Competing interest
SD, FC, and IC are consultants for Thea Laboratories. However, they did not receive any financial support for this study.
Authors’ contributions
SD designed the study, gathered the data, performed the data analysis, and wrote the manuscript. EG, MT, FC, and IC contributed to the data analysis and discussion. All authors read and approved the final manuscript.