Scleritis is a severe and painful ocular disorder, commonly associated with complications that often lead to visual loss and decreased quality of life (1, 2). Approximately half of the cases is associated with a systemic autoimmune mediated diseases, of which rheumatoid arthritis (RA) forms the majority (3). RA classically causes a symmetrical polyarthritis of the small joints (4). The disease may also affect skin, eyes, lungs, heart or nerves, and these extra-articular manifestations considerably increase the burden of disease (5, 6).

Eye involvement in RA usually manifests as scleritis, which is characterized by multiple complications and increased mortality (7). Previous literature showed that scleritis occurred more often in RA patients with autoantibody positivity, including rheumatoid factor (RF) and/or anti-citrullinated protein antibody (ACPA), and erosive disease (8, 9). Most studies on scleritis and RA originate from the pre-biologic era (before the year 2000). In the late 1990’s the first biological agents, such as TNFα-inhibitors became available, which were highly effective in RA. Since the introduction of these biologics the incidence of scleritis in RA might be decreasing according to recent clinical observations (10). However, the exact incidence and severity of scleritis in RA from the current era of modern treatment strategies is not known.

We performed a retrospective cohort study and reviewed the literature on prevalence of scleritis in RA to investigate whether the incidence of scleritis in RA has changed over time as a result of improved treatment options. We also examined the clinical and serological characteristics of recent scleritis cases in patients with RA, who were diagnosed in our tertiary care center.

We performed a retrospective study of all 1623 consecutive patients diagnosed with RA at the department of rheumatology at the Erasmus MC (University Medical Center, Rotterdam, The Netherlands) between 2011 and 2021. The medical files were reviewed for the terms ‘scleritis’ or ‘sclerouveitis’ to asses the incidence of scleritis in RA.

An ophthalmological assessment of all patients with scleritis and RA, who consulted the department of ophthalmology from 2011 to 2021 was also performed (includes also patients treated by rheumatologists from other centers). All had an ophthalmological follow-up of at least 9 months. Included were solely patients with non-infectious scleritis. Collected data included patient demographics, clinical manifestations, complications, and treatment regimen for both scleritis and RA. In addition, diverse laboratory data on the presence of serum autoantibodies, including RF, ACPA and the more recently identified anti-RA33 antibody, during an arbitrary moment in the disease course was reviewed (11). RF (IgA and IgM), ACPA (IgA and IgG), and anti-RA33 (IgA, IgG and IgM) were determined using fluorescence enzyme immunoassay (FEIA) on the Phadia 250 system according to manufacturer’s instructions (Thermo Fisher Scientific, Freiburg, Germany). Whenever possible, missing laboratory data were supplemented from frozen stored sera of included patients.

We reviewed the available literature to evaluate the prevalence and possible risk factors of scleritis in RA. The PubMed database was searched using the terms ‘Scleritis’ and ‘Rheumatoid arthritis’ for all accessible studies up to August 2021. We included studies that fulfilled the following criteria: full text available in English, cohort studies with data on incidence/prevalence, or data on clinical characteristics and complications of scleritis in RA. If available, serological characteristics were also included. Out of 232 available studies, we included 7 studies with data on incidence and/or prevalence as well as reporting on clinical characteristics. Biologicals became commonly available in our country from the early 2000’s, therefore our study was considered to take place in a biologic era. Pre-biologic era was considered for studies in which biological drugs were not yet available.

Continuous data were reported as mean ± range or standard deviation (SD), and categorical data were reported as number with percentage. The local Medical Ethics Committee (Erasmus MC, MEC-2012-016) has reviewed and approved this study, and waived requirement for informed consent. The research was performed according to the Tenets of the Declaration of Helsinki.

Scleritis is one of the severe extra-articular manifestations that can complicate the course of RA. In 1717, William Read described for the first-time ocular disease, later known as scleritis (12). In 1830, Mackenzie attributed scleritis to atmospheric conditions, also known as miasma. He already named it: ‘Scleriotitis atmospherica or rheumatic ophthalmia’. In 1893 the first case of scleral ulceration in a RA patient was described by Holthause (18). In the years thereafter, the relation between RA and (necrotizing) scleritis was well described in several case series, case-control studies, and a few cohort studies (12, 14, 19, 20). Recent clinical observations suggested a decreasing incidence and severity of scleritis in RA, which was attributed to the use of modern therapies, including biologics such as TNFα-inhibitors (3, 21). Our current study shows a cumulative incidence of 0,25% in 10 years, which is indeed to some extent lower than in previous studies.

The disease burden and severity of scleritis in patients with RA were described in detail by Foster et al. in 1984, who reported a high rate of patients with progressive scleromalacia, and a high mortality rate in a cohort before the use of cyclophosphamide (7). Despite the widened current treatment options, our results on scleritis severity are consistent with pre-biologic era reports. The high percentage of patients developing necrosis, even in the current era is striking, though this might have also been influenced by the referral pattern to our ophthalmological tertiary care center.

In our case series, 3 out of 10 patients developed scleritis before the diagnosis of RA was made. This phenomenon was not commonly noted in earlier series, which is probably due to the fact that previous study designs were typically based on inclusion of patients with RA from rheumatologic departments (12, 15, 16). The mean age at onset of RA as well as intervals from RA onset to development of scleritis remained unchanged. Noteworthy is, that all of our patients who developed RA first, had received diverse treatments, including biologics, specifically abatacept, etanercept and adalimumab, before the onset of scleritis, albeit in 3 out of 7 this treatment stopped before scleritis manifested ( Table 3 ). This finding indicates that the withdrawal of immunosuppressive therapy may constitute a risk factor for development of scleritis. Interestingly, none of the patients with RA had ever used rituximab before the onset of scleritis.

The recent report on lower incidence of scleral necrosis (51% versus 80% observed by us), found by Caimmi et al. contrasts with our findings and could be explained by the different inclusion criteria. The study by Caimmi et al. includes a rheumatologic population and possibly provides a better representation of overall scleritis severity in RA. The severity of scleritis may be overestimated in studies that included a tertiary ophthalmological population, which was also the case in 3 of our 10 patients (7, 15). Inflammatory parameters in patients with scleritis and RA vary widely between individual studies. In addition, it is not clear at what moment during the course of the disease the samples were taken and evaluated (12, 14). Caimmi et al. reported that the use of immunosuppressive therapy for RA prior to diagnosis of scleritis was common, which could also interfere with the level of inflammatory parameters (16). In our case series, the laboratory assessment was not standardized and took place at different moments during the disease course. In our view, the most informative are data from the active period of the disease prior to any systemic treatment, but such data were (in our series from tertiary center) not available for all. We consider the measurements of general systemic inflammatory markers from not standardized points of time not consistent for evaluation and/or comparison between the specific groups of patients and therefore these were not included in the present series.

Multiple studies noted that scleritis usually occurs in patients with autoantibody positive RA (presence of RF and/or ACPA) (9, 15, 16). We confirm the high prevalence of erosive disease and autoantibody positivity in patients with scleritis and RA (9, 15, 16). Highly specific serological markers for scleritis in RA are not known. Candidate biomarkers for extra-articular involvement in RA were anti-Ro/SSA antibodies (an anti-extractable nuclear antigen antibody; ENA), and a high level of circulating immune complexes (22, 23). The prevalence of IgA-ACPA (43% vs 34% in the general RA population), and IgA-RF (57% vs 51% in the general RA population) in our patients with scleritis and RA did not differ significantly from the general RA population (24). More recently, other RF and ACPA antibody isotypes besides IgM-RF and IgG-ACPA have been tested in RA, as well as the more recently identified anti-RA33 antibody (24). A combination of positive antibodies could better predict the development of RA, and high titers of all isotypes of RF are associated with worse prognosis of RA. It has also been reported that high IgA-RF levels before DMARD initiation are associated with poor treatment response to TNFα-inhibitors (24). In our small sample these antibodies and their relationship with clinical signs of scleritis could not be evaluated.

Limitations of our study include a small number of patients with scleritis and RA, and its retrospective design. Due to the possibility of retrospective study bias we have not included mild ophthalmic problems such as dry eyes in the present study. Further, it would be valuable to study whether a specific treatment modality such a rituximab could protect patients with RA from scleritis development. For this purpose, a large prospective study would be required and a multicenter and/or international collaboration might be necessary.

To conclude, we found a slightly lower incidence of scleritis in the RA population from the last decade compared to the pre-biological era. On the other hand, we were not able to identify a changing clinical pattern of scleritis in the RA population over time. The majority of patients with scleritis and RA developed severe complications and suffered from scleral necrosis. In addition, the mortality of patients with RA affected by scleritis remained high. Rheumatologists and ophthalmologists should remain aware of these severe scleritis symptoms as well as its association with increased mortality.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Medical Ethics Review Committee Erasmus MC, Erasmus MC, Rotterdam, the Netherlands.

Concept and study design: DV, AR, PJ. Data analysis and interpretation: DV, AR, PJ, MS, JT. Drafting the article: DV, AR, PJ. Critical revision of the article: DV, AR, PJ, MS, JT. All Authors contributed to the article and approved the submitted version.