Disease activity and patient-reported outcomes in patients with rheumatoid arthritis and Sjögren’s syndrome enrolled in a large observational US registry

Sjögren’s syndrome (SS) is a systemic autoimmune disease that can occur independently or in conjunction with another autoimmune condition, such as rheumatoid arthritis (RA) [1, 2]. SS has traditionally been thought of as either primary (SS only) or secondary; however, the terminology for secondary SS has recently evolved to be more descriptive, particularly because SS and other autoimmune diseases are co-existing conditions, rather than one being secondary to the other [2].

The current American College of Rheumatology (ACR) and European League Against Rheumatism (EULAR) treatment guidelines for RA and the ACR/EULAR classification criteria for SS do not include recommendations for treating patients with both RA and SS [3‐5]. However, patients with both RA and SS have an increased disease burden [6, 7] and a decreased quality of life [8] compared with patients with only one autoimmune disease. For example, patients with RA and SS are more likely to have a longer duration of RA and worse joint damage compared with patients with RA only [6, 9]. Patients with SS alone are more likely to experience a negative impact on physical functioning (e.g., lifting and carrying, climbing stairs, bending and walking moderate distances) and social functioning (e.g., quality and quantity of social activities with others) than control subjects without dry eye disease or SS [8, 10].

The correlation of SS with increased disease burden in patients with RA includes an association with higher RA disease activity and anti-citrullinated protein antibody (ACPA) positivity (+) [6, 11]. ACPAs play a pivotal role in the progression of RA [11, 12], indicating a poor prognosis [4], more severe disease course and radiological destruction compared with patients with RA who are ACPA negative (−) [13]. However, there are limited data on the impact of ACPA positivity and SS status in patients with RA and SS [6, 9, 14, 15].

Evaluating real-world data from patients with RA and SS and patients with RA only may enable clinicians to understand the different needs of these populations. The primary objective of this study was to compare RA disease activity and patient-reported outcomes (PROs) in patients from a national registry sample of patients with RA, with and without SS. A secondary objective was to compare RA disease activity and PROs in a sub-group of ACPA+ patients with RA, as measured by anti-cyclic citrullinated peptide (anti-CCP), with and without SS.

SS data were available for 36,256 patients with RA, of whom 1901 met the inclusion criteria (Fig. 1). There were 454 pairs of patients randomly generated with either RA and SS or RA only by frequency matching based on RA duration (Fig. 1 and Supplementary Table S1). Due to residual cohort imbalance after frequency matching (Supplementary appendix), the frequency-matched cohorts were pooled and further refined by PSM. Of the 454 pairs of frequency-matched patients, 283 pairs of patients were PSM (Table 1 and Fig. 1). The baseline demographics, disease characteristics and PROs of PSM patients with RA, with and without SS, were generally well balanced. At baseline, 62–64% of patients were being treated with tumor necrosis factor inhibitors, 21–22% with other b/tsDMARDs and 16–17% with abatacept. At baseline, 56–59% of patients in each cohort were receiving methotrexate monotherapy.

In this analysis of data from the large US Corrona RA registry, we compared RA disease activity and PROs among patients with RA, with and without SS, who initiated b/tsDMARD treatment. Initially, patients with RA and SS or RA only were frequency matched by duration of RA. Significantly fewer patients with RA and SS were in full-time employment and more were disabled or retired compared with patients with RA only. Patients with RA and SS were also more likely to have co-morbidities and to be taking several treatments for RA. These results complement published data examining the effect of SS on quality of life [8, 19]. The symptoms of SS exert a burden on a patient that may impact physical and social functioning, which in turn may affect employment status. Combined with the burden of RA [20], SS may substantially affect a patient’s quality of life.

There is a paucity of published data investigating disease activity in patients with RA and SS, particularly from studies measuring cumulative disease activity with index dates defined by SS onset [6, 9, 15, 21‐23]; only two of the aforementioned studies included a follow-up period [6, 9]. The studies vary in size (82–1471 patients) [9, 22] and methods used for SS diagnosis, often utilizing a mixture of objective (e.g., Schirmer’s test) and subjective methods (e.g., physician exam and/or questionnaire from patient-reported symptoms) [21‐23]. Two cross-sectional observational studies found no relationship between SS status and RA disease activity, as measured by Disease Activity Score 28 (DAS28) using erythrocyte sedimentation rate or C-reactive protein, in patients with RA with and without SS [22, 23]. It should be noted that the proportion of patients with RA and SS in these studies was small: n = 20/82 [22] and n = 11/307 [23]. In contrast, several studies have noted an association between SS and RA disease activity [9, 15, 21]. The analysis of 1471 patients with RA enrolled in a longitudinal RA registry revealed that patients with RA and SS (n = 415) experienced a longer duration of disease, higher RA disease activity (measured by RA disease activity index, CDAI and DAS28) at baseline and significantly lower reduction in RA disease activity at 12 months than patients with RA only [9]. In a study of 636 patients with RA (n = 232 patients with SS symptoms), DAS28 was significantly higher for patients with moderate and severe symptoms compared with patients with no symptoms of SS [21]. Similarly, among 509 patients with RA (n = 74 patients with RA and SS), patients with RA and SS had significantly higher DAS28 scores than patients with RA only [15].

To minimize selection bias in the current study, patients were PSM to account for variables associated with treatment response. The resulting cohorts (patients with RA and SS and patients with RA only) were generally well balanced in terms of baseline characteristics. At the 12-month follow-up visit, there were no significant differences in the change in clinical disease activity between patients with RA and SS and those with RA only; however, change in CDAI score after 12 months was numerically lower (suggesting greater impairment) in patients with RA and SS compared to those with RA only. The lack of association between SS status and RA disease activity in the present study is likely due to PSM; characteristics that are commonly seen with SS, and that are associated with RA disease activity, were accounted for in this analysis, but may not have been in the aforementioned studies, which did not use PSM [9, 15, 21].

Previous studies have demonstrated that patients with SS only (as opposed to patients with SS co-existing with another autoimmune condition) experience pain and fatigue [24], as do patients with RA only [25]; therefore, it could reasonably be inferred that, when co-existing with RA, SS exacerbates these PROs. In this analysis, at the 12-month follow-up visit, improvements in the majority of RA-related PROs were reduced in patients with RA and SS versus patients with RA only (suggesting more impairment). In accordance with this, several studies have demonstrated the additional burden of SS on patients with RA [9, 21, 23]. In an observational study, patients with RA with moderate or severe SS symptoms had significantly worse scores for mHAQ, pain and fatigue compared with patients with RA with no SS symptoms [21]. In a longitudinal registry analysis, patients with RA and SS experienced significantly higher multidimensional (MD) HAQ fatigue scores at baseline compared with patients with RA only [9]. After 12 months, patients with RA and SS experienced numerically lower reductions in MDHAQ fatigue scores than patients with RA only [9]. In addition, an observational study demonstrated that patients with RA and SS experienced numerically greater pain (visual analog scale) scores compared with patients with RA only [23]. Interestingly in our analysis, the differences in two PRO measures, mHAQ and EQ-5D index, between patients with RA and SS and patients with RA only, were marginal. However, the mHAQ and EQ-5D index questionnaires are completed by patients and would likely be impacted by underlying RA disease status.

Patients with RA who are ACPA+ are more likely to have a more severe disease course than patients who are ACPA− [13]. A study of Greek patients with RA concluded that the presence of anti-CCP antibodies was associated with extra-articular manifestations, such as serositis and pulmonary fibrosis [14]. In several studies, the proportion of patients with anti-CCP antibodies has been found to be higher in patients with RA and SS compared with patients with RA only (67.6–77.8% versus 59.4–71.9%) [6, 9, 15]. Therefore, we sought to clarify if the presence of SS would exacerbate the disease course further in anti-CCP+ patients. Contrary to previous research [6, 9, 15], in the current study, the proportion of patients with anti-CCP antibodies was lower in patients with RA and SS compared with patients with RA only (49.1% versus 55.5%); however, it should be noted that not all patients had measures reported. RA disease activity and PRO results were similar in the subgroup of patients with RA who were anti-CCP+, indicating that SS status did not affect outcomes differently in this patient population.

The biologic pathways associated with the differences in patients with RA with, and without, dry eyes and mouth (SS) are not clear [26‐28]. However, the results of the current study do suggest that physicians may wish to consider SS status in the management of patients with RA; in particular, there may be a need for closer monitoring and assessment of response to treatment. More aggressive treatment of SS with RA may be needed.

This analysis has several strengths, namely the Corrona registry is the largest disease registry in the US that collects data directly from both providers and patients at the time of a routine clinical encounter. This allowed for patients with RA and SS or RA only to be selected from a broad population, ensuring satisfactory pairing [29]. The Corrona registry does include a wide variety of rheumatology practices participating throughout the country (rural and urban areas, academic and private settings) and brings access to broad geographic locations and patients with diverse sociodemographic origins. Prior analyses compared Medicare patients with RA enrolled in Corrona to those who are not part of the registry and found similar demographic and co-morbidity characteristics, supporting the generalizability of the Corrona registry [29]. Data are collected at regular intervals, which enabled us to evaluate outcomes at two different time points. Advanced epidemiological methods (e.g., PSM) were used to compare responses between patients with and without SS where selection bias may have existed. The presence of SS was captured on the provider form and is considered a critical field of data for Corrona. The results from this observational study in US patients complement previous studies [8, 9, 19, 21‐23].

As opposed to previously reported studies, bias was minimized by PSM of the two cohorts (RA and SS versus RA only) by factors known to be associated with treatment response (e.g., age, sex, CDAI score, number of prior biologics, work status, history of co-morbidities and RA disease characteristics). Another limiting factor was that patients included in the registry were diagnosed by different rheumatologists across the US; diagnosis of SS was at the discretion of the treating physician (SS associated with RA: yes/no) and rheumatologists may have used a variety of mostly historical clinical signs and symptoms rather than objective testing (i.e., Schirmer’s test). Nevertheless, we believe that the manner in which these data were collected reflects real-world clinical practice with the reporting of dry eyes and/or mouth not associated with medications or mouth breathing during the night. Additionally, patients in the RA-only cohort also required a questionnaire entry for SS associated with RA (yes/no), potentially leading to under-ascertainment.

In this large US patient population initiating b/tsDMARD treatment, patients with RA only had greater improvements in RA disease activity and PROs than those with RA and SS; similar results were shown in anti-CCP+ patients. Physicians may wish to consider SS status when managing patients with RA; specifically, patients with RA and SS may require closer monitoring and more aggressive intervention to improve their disease experience than patients with RA only. Additional studies are needed to further understand the biological pathways involved in SS co-existing with RA, and subsequently to provide targeted treatment options for this population of patients.