Scleritis

Scleritis is defined as inflammation of the opaque outer eye wall, or sclera, and is classified most commonly as either anterior or posterior.1–4 Anterior scleritis is visible on direct clinical inspection of the exposed eye wall and is subdivided further as diffuse, nodular, or necrotizing. In contrast, inflammation of the posterior sclera is not visible directly and so requires ultrasonographic demonstration of posterior eye wall thickening, often in the presence of retrobulbar and perineural fluid producing a pathognomonic ‘T-sign.’ Whereas posterior scleritis can also be diffuse or nodular, necrotizing posterior scleritis is not generally recognized clinically. The symptoms of pain and/or headache are reported frequently by patients with scleritis and are often worse at night due to dependent or positional tissue swelling. The occurrence of an associated autoimmune disease and the need for systemic corticosteroid and/or non-corticosteroid immunosuppressive therapy are common, particularly in patients with anterior nodular or necrotizing disease, or with posterior segment inflammation. Infectious scleritis is rare, but can occur in the setting of local infection, such as herpes simplex and varicella zoster virus infection; with systemic infections, such as tuberculosis or syphilis; and following eye wall surgery—particularly pterygium removal or scleral buckle placement.5–7 Six original articles8–13 and three letters14–16 in this issue of Ocular Immunology & Inflammation (OII) present important findings relevant to the clinical presentation, diagnosis, and management of scleritis.

Lavric et al.8 reported the results of a retrospective review of the demographic, clinical, fluorescein angiographic, and ultrasonographic features of 114 patients diagnosed with posterior scleritis over a 10-year period, from 2004 to 2013, at two referral centers: Moorfields Eye Hospital in London, UK (n = 74; 64.9%) and Sankara Nethralaya Medical Research Foundation in Chennai, South India (n = 40; 35.1%). The authors performed a patient-based analysis of pooled data from the two centers. For subjects with bilateral involvement, data from only the more severely affected eye was included in the study. The mean age of the combined cohort was 45.9 years and nearly three-quarters (71.1%) were women. An associated autoimmune condition was identified in 37.7% of patients, the most common of which were rheumatic polyangiitis (n = 14; 12.3%), systemic vasculitis associated with pANCA positivity (n = 6; 5.3%), and systemic lupus erythematosus (n = 5; 4.4%). Infectious causes were rare, and included one patient each (0.9%) with evidence of Lyme disease, varicella zoster infection, and congenital rubella syndrome. Vision at presentation was worse than 6/12 in over two-thirds of patients and periocular pain was the most common symptom (64.0%), followed by blurred vision (34.2%) and headache (13.2%). Common anterior segment findings included conjunctival chemosis (31.6%), conjunctival hyperemia (29.8%), anterior chamber cells (25.4%), and evidence of anterior scleritis (19.3%). Common posterior segment findings included choroidal folds (28.1%), serous retinal detachment (28.1%), macular edema (22.8%), optic disc edema (14.0%), vitreous cells (11.4%), and a subretinal mass (7.0%). B-scan ultrasonography revealed a posterior eye wall thickness >2.0 mm in 54.4% and the presence a T-sign in 41.2%. Treatment was generally similar for subjects with or without an associated systemic disease, and included various combinations of topical corticosteroids (45.6%), oral corticosteroids (38.6%), oral non-steroidal anti-inflammatory drugs (NSAIDs) (34.2%), and a systemic non-corticosteroid immunosuppressive agent (32.5%). Remission was defined as inactivity off both corticosteroids and non-corticosteroid immunosuppressive agents for at least 90 days. A total of 101 subjects (88.6%) were both controlled with treatment and followed for this minimal 3-month period, with a median time to remission of approximately 7 months. Just over one-third of the patients (36.6%) experienced at least one recurrence, with a relapse rate of 15.8% per person-year. While the numbers were quite small, both the diagnosis of Crohn disease (n = 3; hazard ratio (HR) 12.3; nominal p value = 0.018) and treatment with mycophenolate mofetil (n = 8; HR 2.98; nominal p value = 0.041) were associated with increased risk of recurrence. Mean vision improved between presentation and last visit, by −0.24 ± SD 0.36 logMAR (∼2–3 lines Snellen equivalent; nominal p value <0.001). While patients seen in India tended to be slightly younger (40.5 vs 48.8 years; nominal p value 0.004) and to have slightly worse vision at presentation (logMAR 0.60 (Snellen ∼20/80) vs 0.29 (Snellen ∼20/40); nominal p value 0.001), there were otherwise no statistically significant differences in the clinical presentation, associated systemic disease, time to relapse, or complications in the patient cohorts from the two centers. Vision at last visit was comparable in the two subgroups (logMAR 0.19 in India vs 0.15 in the UK; nominal p value 0.56; logMAR 0.18 = Snellen 20/30). A comparison by the authors of their results to findings in four previously reported, large, retrospective clinic-based cohort studies, suggested a similar spectrum of symptoms and signs, but highlighted the relatively wide variation in clinical presentation across both patients and centers.

In a companion paper of the same 114 subjects from the two referral centers in India and the UK, González-López et al.9 compared the subgroup of 18 patients (15.8%) with bilateral posterior scleritis to the much larger subgroup with unilateral disease. Again, for those with bilateral inflammation, the more severely affected eye as defined by worse visual acuity was selected for study, and for all patients, remission was defined as above. Generally speaking, with the exception of perhaps more headache as opposed to periocular pain in subjects with bilateral inflammation, the authors identified few differences in the demographic, clinical or ultrasonographic features, association with systemic conditions, approach or response to treatment, or recurrence or relapse rates in the two subgroups. In addition to obvious genetic and socioeconomic differences between patients from India and the UK, the authors acknowledged clear center-specific differences in treatment and the possibility for differential treatment bias. Other sources of bias, such as referral, selection and classification bias, together with the decision to exclude the findings of the 18 fellow eyes with less severe involvement from the descriptive and univariate analyses, no doubt also influenced the results.

Somkijrungroj et al.10 investigated conjunctival sensation in nine patients with active and 19 patients with inactive anterior scleritis, including 11 patients with an associated autoimmune disorder, seven of which were bilateral. A total of five eyes had infectious scleritis, including three due to varicella zoster virus in the setting of herpes zoster ophthalmicus, and one each associated with local infection due to herpes simplex virus and Mycobacterium chelonae. A single examiner used the same hand-held, Cochet-Bonnet esthesiometer to quantify sensation 5 mm from the limbus in all areas of active and inactive scleritis, as well as the horizontal medial and temporal and vertical inferior and superior quadrants of both eyes, whether affected or not. A 7-point scoring scale was applied to each measurement, wherein zero (0) was assigned for no conjunctival sensation and six (6) a very sensitive conjunctiva. Corresponding areas in the contralateral eye were used as controls. Mean sensitivity measurements were generally modest and <2.5 in both affected and unaffected eyes and areas, regardless of disease activity. The authors identified no change in conjunctival sensation over areas of active, non-infectious scleritis—although the sample size was small at nine subjects. Modest mean decreases in sensation at or slightly above a single Cochet-Bonnet level were identified in areas of previously active non-infectious scleritis (nominal p value <0.001) and in all areas of eyes with inactive infectious scleritis—including both uninvolved areas and areas with previously active scleritis. The authors suggested that formalized testing of conjunctival sensation might aid in the diagnosis of anterior scleritis, particularly, by inference, in eyes suspected of having infectious scleritis due to herpes zoster infection, where conjunctival sensation in the affected eye tended to be more deeply and diffusely affected. The authors noted the potential for bias in their retrospectively studied referral cohort and cautioned against confounding when testing and interpreting sensation clinically—particularly in patients on chronic topical medications, which can independently reduce sensation.

Axmann et al.11 described the use of enhanced depth anterior segment optical coherence tomography (AS-OCT) to study 11 consecutive, prospectively identified subjects. Isolated diagnoses included nodular scleritis (n = 4), diffuse scleritis (n = 3), nodular episcleritis (n = 2), and diffuse episcleritis (n = 3). The authors demonstrated that the average anterior eye wall thickness in unaffected eyes was approximately 800 μm, that this thickness increased by about 200 μm in eyes with both episcleritis and scleritis, and that vascular dilatation tended to be more superficial in eyes with episcleritis. Use of AS-OCT revealed scleral thinning in one eye that had recurring bouts of scleritis over the period of the study. The authors suggested that enhanced depth AS-OCT might be used to both diagnose and monitor scleritis and episcleritis.

Agrawal et al.12 retrospectively studied the use of oral flurbiprofen (50 or 100 mg three times daily) to treat 126 patients with non-infectious, non-necrotizing nodular or diffuse anterior scleritis. Sixty-one of these patients (48.4%) were also treated with topical corticosteroids, whereas the remaining 65 (51.6%) were not. The authors stated that they tended to use oral flurbiprofen alone to treat eyes with mild scleritis and oral flurbiprofen plus a topical corticosteroid in eyes with mild to moderate scleritis. Eyes that failed to respond to combined treatment were escalated to either oral corticosteroids or a systemic non-corticosteroid immunosuppressive agent. The two treatment cohorts were generally balanced with regard to demographics with a mean age of about 53 years and women constituting two-thirds of patients. The scleritis was bilateral in about 10% of subjects and just over three-quarters of affected eyes had diffuse anterior scleritis. Eighteen (14.3%) of subjects had an underlying autoimmune condition. All but 10 patients (92.1%) were controlled on one of these two treatment regimens, an overall observed failure rate of 1.07 (95% confidence intervals (CI): 0.57–1.99) per 1000 person-years. Treatment failures were, however, 10-fold higher (HR 95% CI: 2.52–39.65) in patients with an associated autoimmune condition. The average duration of treatment was about 6 weeks, although treatment times varied widely from ≤2 weeks to over a year. Possible treatment-related adverse events included gastrointestinal irritation in 12 patients (9.5%), nausea, headache, and cataract in two patients each (1.6%), and skin rash in one patient (0.8%). The authors concluded that oral NSAIDs either alone or in combination with topical corticosteroids were highly effective and generally very well tolerated for the treatment of mild–moderate anterior scleritis, but caution that such therapies were often inadequate to achieve control in those patients with an underlying autoimmune condition. Both Bauer et al.17 and Sainz de la Maza et al.18 reported comparably high rates of control of mild– moderate non-infectious, non-necrotizing anterior scleritis using an oral NSAID as first-line therapy.

Yan and Zhang13 described a 43-year-old Chinese woman who developed painless posterior scleritis in the setting of immunoglobulin-A nephropathy (IgAN), and immune complex-mediated glomerulonephritis. Findings included serous retinal detachments, multiple retinal pigment epithelium (RPE) leaks on fluorescein angiography, and posterior eye wall thickening with a ‘T-sign’ on ultrasonography in each eye. A review of the literature performed by the authors identified one additional case of IgAN-associated posterior scleritis, four cases of anterior scleritis, and 12 cases of episcleritis. Patients were treated variably with oral flurbiprofen with or without systemic and topical corticosteroids, or systemic non-corticosteroid immunosuppressive therapy.

Ali and Young14 reported a 59-year-old man with progressive posterior scleritis in the setting of necrobiotic xanthogranuloma (NXG) that was refractory to treatment, similar to the case mentioned by Lavric et al.8 in this issue of OII. The diagnosis was made following biopsy of a suspicious skin lesion. The authors cite several additional cases of scleritis associated with NXG and note that the inflammation often fails to respond to systemic corticosteroids alone. They mention that yellow-orange skin papules and plaques are common in the disorder, and that up to 80% of patients will have an associated paraproteinemia that may progress to hematologic malignancy.

Fénolland et al.15 described a 47-year-old man who developed nodular anterior scleritis with underlying peripheral retinitis, vasculitis and vitritis as the presenting signs of syphilis. Serologic testing revealed co-infection by the human immunodeficiency virus (HIV). The scleritis resolved following intravenous penicillin G therapy. A number of previous studies have described syphilis scleritis—often nodular.19–22

Schuler et al.16 reported a 47-year-old woman with long-standing rheumatoid arthritis and recurring nodular anterior scleritis. Posterior scleritis with macular edema was also present. Scleral necrosis developed and the patient’s systemic immunosuppressive therapy was modified over time to include prednisone, adalimumab, cyclophosphamide, and finally rituximab—which controlled the inflammation. While on rituximab, the patient developed necrotizing herpetic retinitis diagnosed clinically as acute retinal necrosis and confirmed by polymerase chain reaction-based testing of anterior chamber fluid to be due to herpes simplex virus type 1. The retinitis responded promptly to intravenous acyclovir and oral prednisone. Takei et al.23 described a patient with malignant systemic lymphoma who developed necrotizing herpetic retinitis following treatment with rituximab.

Together, these studies highlight the broad clinical spectrum of scleritis. They also emphasize the need for a high index of suspicion, both with regard to initial diagnosis in eyes with posterior scleritis, and when evaluating patients, be it for an associated autoimmune condition or when considering the possibility of a local and systemic infection – which may be either causal or a consequence of systemic immunosuppressive therapy. For non-infectious scleritis, treatment should be based on severity. The vast majority of patients with mild– moderate non-infectious anterior scleritis will respond to an oral NSAID with or without a potent topical corticosteroid, such as dexamethasone, prednisolone acetate, or difluprednate. More aggressive treatments, including systemic corticosteroids and non-corticosteroid immunosuppressive therapies, should be reserved for patients who fail such initial therapies, who present with severe nodular or necrotizing disease, or when needed to control systemic inflammation.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Supported in part by The Pacific Vision Foundation (ETC) and The San Francisco Retina Foundation (ETC).