Antineutrophil Cytoplasmic Antibody–Associated Active Scleritis

Objective  To determine whether antineutrophil cytoplasmic antibody (ANCA) testing provides prognostic information in evaluating scleritis.

Methods  Retrospective medical record review of patients evaluated at a tertiary care center from January 1, 1995, to June 30, 2006, was performed to compare clinical features, treatments, and associated systemic disease in patients who test positive for ANCA vs patients whose ANCA tests are negative.

Results  Among 78 patients identified, 14 tested positive for ANCA. Patients with positive ANCA test results were more likely to have an associated systemic disorder (10 of 14 or 71%) than were patients who tested negative for ANCA (26 of 64 or 41%; P = .04), and the disorder was more likely to have been diagnosed as a result of scleritis work-up (2 of 10 or 20% vs 19 of 26 or 73%; P = .007). Patients with positive ANCA test results had significantly more ocular complications (21 of 14 or 86% vs 20 of 64 or 31%; P < .001), including keratopathy (5 of 14 or 36% vs 6 of 64 or 9%; P = .02), visual acuity of 20/50 or worse (8 of 14 or 57% vs 11 of 64 or 17%; P = .001), and vascular pannus (3 of 14 or 21% vs 1 of 64 or 2%; P = .02). Aggressive therapy, such as chronic systemic corticosteroids (9 of 14 or 64% vs 9 of 64 or 14%; P < .001) and alkylator therapy (8 of 14 or 57% vs 7 of 64 or 11%; P < .001), was more likely to be recommended for patients who tested positive for ANCA.

Conclusions  A substantial subset of patients with scleritis are also positive for ANCA. These patients are more likely to have severe ocular disease and undiagnosed primary vasculitic disease, thereby requiring more aggressive therapy. An ANCA test may be useful in the evaluation and treatment of patients with scleritis.

Scleritis is a destructive inflammatory disorder of the outer coating of the eye that usually manifests as redness and severe ocular pain. The inflammation can spread to surrounding structures of the eye, progress to ischemia and necrosis, and potentially cause severe visual loss or blindness.^1-3 Nearly half of patients with scleritis have an associated immune-mediated condition, such as rheumatoid arthritis (RA), relapsing polychondritis, inflammatory bowel disease, or primary vasculitis.²

Among patients with an associated disease, most patients seek treatment for systemic disease before the onset of scleritis.² However, patients with a primary vasculitic disease, such as Wegener granulomatosis (WG) or polyarteritis nodosa (PAN), have a more aggressive form of scleritis and are more likely to first seek treatment for scleritis.^2,3 If left untreated, primary vasculitic disease leads to multiorgan failure and is potentially fatal.^3,4 The combination of corticosteroid and cyclophosphamide treatment has been shown to improve survival in patients with WG and PAN.^3,4 Therefore, early diagnosis and initiation of treatment are critical for patients with scleritis associated with primary vasculitic diseases.

Diagnosing systemic vasculitides may be difficult and often relies on a combination of clinical, pathological, and immunological criteria. In clinical settings, when patients have active inflammation and high pretest probability, the detection of antineutrophil cytoplasmic antibodies (ANCA) serves as a useful clinical marker for vasculitis.^5,6 With an indirect immunofluorescence assay, ANCA stains in 2 major patterns. The cytoplasmic pattern (c-ANCA) detects a neutrophilserine proteinase and exhibits a diffuse staining throughout the cytoplasm. Alternatively, the perinuclear staining pattern(p-ANCA) is exhibited when antibodies detect lysosomal enzymes such as myeloperoxidase. Typically, c-ANCA has been associated with WG, and p-ANCA has been linked to renal vasculitis and microscopic polyarteritis.^2-7 However, ANCA testing is not definitive for systemic vasculitis, and there is overlap in clinical phenotypes among patients with p-ANCA or c-ANCA as well as patients negative for ANCA.

Although ANCA testing plays a critical role in the evaluation of vasculitides, its clinical significance in patients with scleritis has not been clearly established. A previous study reviewed the clinical features of and treatment for scleritis and gave insight into the value of ANCA testing.¹ The study found a subset of patients with scleritis who tested positive for c-ANCA but had no evidence of an associated systemic disease.¹ These patients had difficult-to-control scleritis, suggesting c-ANCA may be an independent prognostic marker.

To investigate the prognostic significance of ANCA testing in scleritis, we evaluated patients who tested negative for ANCA and compared this group with patients who tested positive for p-ANCA or c-ANCA. We also examined ocular complications, treatments, associated diseases, and timing of the systemic diagnosis relative to scleritis.

Medical records from all patients with scleritis evaluated at the Inflammatory Eye Disease Clinic, Casey Eye Institute, Oregon Health & Science University between January 1, 1995, and June 30, 2006, were retrospectively reviewed. Patients seen before 1995 were excluded because ANCA screening was not routinely administered. The study protocol was reviewed and approved by the Oregon Health & Science University Institutional Review Board.

Scleritis was diagnosed based on characteristic clinical findings of inflammation and edema affecting the scleral tissues, with involvement of the superficial and deep episcleral vascular plexus.⁸ Posterior scleritis was confirmed by B-scan ultrasonography, computed tomography, and/or magnetic resonance imaging of the orbits.^7,8 Scleritis was classified as diffuse anterior, nodular, necrotizing, scleromalacia perforans, or posterior based on the system devised by Watson and Hayreh.⁸

Patient data, including age, sex, eye involvement, and follow-up period, were recorded. Ocular complications included visual impairment, scleral thinning, corneal involvement, uveitis, elevated intraocular pressure, and posterior segment involvement. Visual impairment was categorized as visual acuity of 20/50 or worse and visual acuity of 20/200 or worse. Scleral thinning was defined as degeneration of the sclera leading to loss of scleral rigidity.⁷ Corneal involvement included any signs of corneal infiltrates, corneal thinning, stromal keratitis, and corneal ulceration.⁷ Uveitis was classified according to the guidelines from the Standardization of Uveitis Nomenclature Working Group.⁹ Elevated intraocular pressure was defined as greater than 21 mm Hg. Cystoid macular edema was defined by clinical and angiographic criteria. Posterior segment involvement included any findings of optic nerve edema, serous retinal detachment, or choroidal folds.⁷ Because the percentages of complications found are potentially affected by the duration of follow-up, we further analyzed the rate at which these outcomes occurred.¹⁰

Associated systemic disease was diagnosed based on compatible medical history, physical findings, imaging, laboratory results, and confirmation by an appropriate specialist. All patients were evaluated by a rheumatologist. At the initial evaluation of scleritis, nearly all patients were routinely screened for ANCA immunofluorescence to determine the presence of c-ANCA or p-ANCA. Results of the screening were confirmed by radioimmunoassay for the presence of antibodies to serine proteinase 3 or myeloperoxidase. All tests ordered from January 1, 1995, through June 30, 2006, were performed by ARUP Laboratories (Salt Lake City, Utah) using indirect fluorescent antibody/multianalyte fluorescent detection. Patients with exclusively infectious causes of scleritis or episcleritis were not screened for ANCA and were excluded from the study. Patients with inactive scleritis were considered to have low pretest probability and were also excluded. Patients who had a clearly established connective tissue disorder such as RA and who had no evidence of a second associated condition were not screened for ANCA. Other routine tests included urine analysis, serum creatinine, and fluorescent treponemal antibody absorption. Additional tests such as rheumatoid factor, antinuclear antibody, erythrocyte sedimentation rate, C-reactive protein, and chest radiography were performed if deemed appropriate by a rheumatologist, based on the patient's history and physical examination.

Treatment was based on type of scleritis, severity of disease, associated medical conditions, and patient preference. Drug dosage and tapering were determined by clinical response and tolerance of the medication. For most patients with mild anterior scleritis, an oral nonsteroidal anti-inflammatory drug was the first-line treatment. For complicated, refractory, necrotizing, or posterior scleritis, systemic corticosteroid treatment (oral or intravenous) was initiated. Chronic corticosteroid therapy was defined as use for more than 6 months over a 1-year period. Indications for the use of immunosuppressive drugs/steroid-sparing agents included failure of scleritis to respond to corticosteroid therapy, recurrence of scleritis, corticosteroid toxicity, the physician's judgment that a corticosteroid-sparing drug was indicated, or presence of active associated systemic disease. Corticosteroid-sparing medications included antimetabolite therapy (azathioprine, methotrexate sodium, and mycophenolate mofetil), alkylator therapy (chlorambucil and cyclophosphamide), and others (cyclosporine, infliximab, rituximab, and intravenous immunoglobulin).

Clinical characteristics, treatments, and associated medical conditions were compared between patients positive for ANCA and patients negative for ANCA and between patients positive for c-ANCA and p-ANCA. Distributions of categorical variables were compared using the Fisher exact test with exact procedures for the calculation of significance. Continuous variables were compared using the 2-tailed t test. Nominal P values were used for all comparisons, and P < .05 was considered statistically significant.

Of 101 patients with scleritis, 78 met the inclusion criteria, and 23 were excluded because they had inactive disease, episcleritis exclusively, or infectious scleritis. Of the 78 patients included, 14 patients (18%) tested positive for ANCA, and 64 were considered ANCA negative, including 52 patients with a negative ANCA test result and 12 patients with clearly established RA without evidence of a second immune-mediated disorder. Demographic characteristics and scleritis classifications were not significantly different between the 2 study groups (Table 1). Most patients in both subgroups were women aged 40 to 50 years. The mean follow-up for ANCA-positive patients (22 months; range, 0-86 months) was longer than that for ANCA-negative patients (9 months; range, 0-63 months), but this difference was not statistically significant (P = .09). Anterior diffuse scleritis was the most common type in both groups. Bilateral eye involvement was present in approximately half thepatients.

Patients who tested positive for ANCA experienced more complications than patients who tested negative for ANCA (86% vs 31%; P < .001). Visual impairment occurred more frequently in ANCA-positive patients than in ANCA-negative patients (visual acuity ≤20/50, P = .001; visual acuity ≤20/200, P = .39) (Table 1). Keratopathy was 3 times more common in patients positive for ANCA than it was in patients negative for ANCA (P = .02). Vascular pannus also was seen more often in patients positive for ANCA (P = .02). In addition, more patients positive for ANCA had scleral thinning, uveitis, elevated intraocular pressure, and posterior segment complications, but the number of patients involved was too small for statistical analysis.

Outcomes are frequently affected by the duration of follow-up.¹⁰ As noted above, the duration of follow-up differed slightly between groups. To analyze the effect of duration of follow-up, we calculated the rate of development of specific outcomes in terms of patient years. These rates did not differ significantly between patients positive for ANCA and patients negative for ANCA (Table 2).

Most patients were treated with a systemic medication, including nonsteroidal anti-inflammatory drugs (Table 1). Most required systemic corticosteroids, but far more patients positive for ANCA were treated with long-term corticosteroid therapy (P < .001). A steroid-sparing agent was also used with significantly greater frequency among patients positive for ANCA (P = .002). This difference is most striking when comparing the use of alkylator therapy, 57% (8 of 14) of patients positive for ANCA required alkylator therapy compared with 11% (7 of 64) of patients negative for ANCA (P < .001).

Patients positive for ANCA also had systemic disease significantly more often than patients negative for ANCA (P = .04); in particular, ANCA-positive scleritis was associated with a primary systemic vasculitis (P < .001) (Table 1). The most common primary systemic vasculitis associated with ANCA was WG, which was seen in 7 patients (50%), followed by PAN, which was seen in 2 patients (14%). Among patients with WG positive for ANCA, 3 had generalized WG and 4 had limited WG. Limited WG was also diagnosed in 1 patient with a negative ANCA test.

Scleritis was more likely to be the presenting feature of an associated systemic disease in patients positive for ANCA. Of 26 patients negative for ANCA with an associated disease, 19 (73%) presented with a known history of the concurrent systemic disease. In contrast, 2 of 10 patients positive for ANCA with an underlying disease (20%) also had a pre-existing systemic diagnosis (P = .007).

Within the ANCA-positive group, 7 (50%) were positive for c-ANCA, and 7 (50%) for p-ANCA (Table 3). Patients positive for p-ANCA were significantly older at initial treatment for scleritis than were patients positive for c-ANCA (mean age, 64 years [range, 49-84 years] vs 46 years [range, 30-61 years]; P = .01). All patients positive for p-ANCA had anterior diffuse scleritis; conversely, patients positive for c-ANCA had an assortment of scleritis types, including 3 with diffuse, 2 with nodular, and 2 with posterior scleritis. Scleritis type correlated with complications in both subgroups. In patients positive for p-ANCA, 2 had scleral thinning, and 4 had corneal complications; in comparison, 4 patients positive for c-ANCA had scleral thinning, 3 had uveitis, 2 had corneal complications, 2 had vascular pannus, 1 had posterior segment complications, and 1 had elevated intraocular pressure. Visual impairment was present in all patients positive for c-ANCA and 5 of 7 patients positive for p-ANCA. Nearly all patients in each subgroup required a steroid-sparing agent in addition to an oral corticosteroid. The use of alkylators and antimetabolites was similar in both subgroups. Association with a primary vasculitis was also similar in both groups. Among patients positive for c-ANCA, 5 had WG, and 1 had Cogan syndrome. Among patients positive for p-ANCA, 2 had WG and 2 had PAN.

We examined the prognostic value of ANCA testing in scleritis. Our results indicated that ANCA testing identifies a subset of patients with scleritis who are more likely to have severe ocular disease, require multiple immunosuppressive therapies, and have an underlying vasculitis diagnosed as a result of their scleritis. We also investigated the difference between p-ANCA scleritis and c-ANCA scleritis. Patients positive for p-ANCA had diffuse anterior-type scleritis, whereas patients positive for c-ANCA had heterogeneous types of scleritis; otherwise, there was no striking distinction between c-ANCA and p-ANCA scleritis. Because there was a small number of patients in both subgroups, statistical analysis was limited, and larger studies are required for futureinvestigations.

Most patients with scleritis positive for ANCA presented de novo without any history of an immune-mediated condition. With further work-up, most of these patients were found to have an underlying primary vasculitic disease, most commonly WG. This correlates with the results found in a previous study, which showed that patients with a primary vasculitis are more likely to be identified as a consequence of their scleritis.² Because of the potential for life-threatening complications, all patients with scleritis should have an ANCA test performed, and those who test positive should be thoroughly evaluated for an underlying systemic vasculitis. In this respect, we use the ANCA test in evaluating patients with scleritis far differently than the rheumatoid factor or antinuclear antibody tests. We do not routinely screen patients with scleritis and a rheumatoid factor or antinuclear antibody because (1) scleritis is rarely the presenting manifestation for either RA or systemic lupus erythematosus; and (2) the specificities of rheumatoid factor and antinuclear antibody are such that a diagnosis of either RA or systemic lupus erythematosus cannot be established by laboratory values in the absence of a compatible clinical presentation. In contrast, a positive ANCA value can identify a population that is at increased risk for developing a systemic vasculitis, which has a substantially worse prognosis than other causes of scleritis. We do not recommend repeating the ANCA test once it is found to be negative unless the patient has developed additional evidence of systemic disease, such as glomerulonephritis.¹¹

Significantly more patients positive for ANCA had long-term prescriptions for systemic corticosteroids and alkylator therapy than patients negative for ANCA. The more aggressive treatment required for ANCA-positive scleritis is confounded by the associated immune condition in most patients positive for ANCA. As was shown in previous studies, patients with scleritis and a primary systemic vasculitis, such as WG or PAN, often require systemic corticosteroids combined with alkylator therapy for control of their eye disease.⁴ There may also be a bias to initiate more aggressive treatment because of the poor prognosis of patients with primary vasculitic diseases.

Patients positive for ANCA are at increased risk for developing visual impairment, keratopathy, and vascular pannus. Corneal changes from corneal infiltrates, thinning, or ulceration were the most common threat to vision. Although patients with ANCA-positive scleritis had significantly more ocular complications, a referral bias may exist because our clinic is part of a tertiary care medical center and is likely to see more severe cases of scleritis.

As with all retrospective studies, our data must be interpreted with caution. We found a significant subset of patients with scleritis who were positive for ANCA. Although ANCA test results may serve the physician as adjunct evidence for the diagnosis of vasculitis, these results must be viewed in the context of the patient's clinical picture and disease activity. Testing for ANCA is most valuable in settings where there is a high pretest probability of vasculitis. We do not routinely screen patients with inactive scleritis and no clinical evidence of an associated systemic disorder.

The ANCA test has become so closely associated with the diagnosis of WG that some authorities use it as a diagnostic criterion. This would effectively bias any analysis of the relationship between a test result positive for ANCA and systemic disease. To avoid this bias, the diagnosis of limited WG was based on involvement of multiple organs other than the lungs or kidneys. Typical sites of disease included the eyes, ears, sinuses, nose, or upper airway. These organs do not usually allow a histological confirmation of vasculitis. However, a positive ANCA test was not a requirement to diagnose limited WG, and the diagnosis was applied to patients who were negative for ANCA as well as patients who tested positive.¹²

A positive ANCA test result heralds an underlying primary systemic vasculitis that may not be clinically apparent. Despite the limitations inherent in a retrospective study, we conclude that ANCA testing is a valuable screening tool in evaluating patients with active scleritis. A positive ANCA test result is a poor prognostic factor. Patients with ANCA-associated scleritis tend to have more visual impairment, corneal complications, and vascular pannus, and these patients often require combined chronic systemic corticosteroid and alkylator therapy for control of their eye disease. Further studies would be helpful in distinguishing any differences between patients with scleritis and positive p-ANCA test results vs c-ANCA test results.

Correspondence: Lani T. Hoang, MD, Casey Eye Institute, Oregon Health & Science University, 2934 NW Telshire Terr, Beaverton, OR 97006.

Submitted for Publication: June 1, 2007; final revision received November 12, 2007; accepted November 19, 2007.

Author Contributions: Drs Hoang, Choi, and Rosenbaum had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Financial Disclosure: None reported.

Funding/Support: This study was supported in part by grant 5-P30-EY010572 from the National Institute of Health, a Research to Prevent Blindness Senior Scientific Investigator Award, and the Stan and Madelle Rosenfeld Family Trust (Dr Rosenbaum).

Role of the Sponsor: The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.