Flares of systemic disease in primary Sjogren's syndrome

Sir, Rheumatic disorders such as rheumatoid arthritis (RA) or systemic lupus erythematosus (SLE) are characterized by a relapsing–remitting course, with flares of disease activity [1, 2] amenable to therapy. This has not been well studied for the systemic features of primary Sjögren's syndrome (PSS). We report our observations from 59 consecutive patients with PSS, fulfilling the revised EU–USA criteria [3], who attended a multidisciplinary Sjögren's syndrome clinic in Birmingham, UK between 1 January 1997 and 31 December 2001. Six further patients were lost to follow-up and one patient who developed SLE was excluded from this analysis. Patients were generally seen on a 6-monthly basis with additional visits if clinically indicated. Patients were asked about their disease activity since they were last seen. Local Research Ethics Committee approval was obtained. For this study, a flare was categorized as any acute worsening of extra-glandular features potentially of inflammatory origin defined in the BILAG index for SLE [2]. The Mann–Whitney U-test was used to compare continuous data and the χ2 test to compare discontinuous data between groups. A P value of less than 0.05 was taken to be significant. Nine patients had a flare with a total of 12 different episodes (i.e. median number of flares per patient = 0 and median number of flares in patients who had one = 1). This represents a crude prevalence of 15% of patients having flares over the 5-yr period of the study (3% per year), or, if presented as the percentage of patients with a flare over the median period of follow-up (27 months), 7% per year. All the patients who had a flare were female (100%) reflecting the cohort generally [48/50 (96%) of those without were also female]. They were slightly younger with a median age of 50.0 yr (range 39–58 yr) compared with 58.5 yr (range 30–84 yr) in those without a flare (P = 0.018). Although the median disease duration (at 31 December 2001) in those with and without a flare was similar [4.0 yr (range 1–15 yr) and 4.5 yr (range 0–31 yr)], the median period of follow-up within the clinic was longer for the flare group [38 months (range 11–58 months) compared with 25 months (range 1–60 months); P = 0.017] (i.e. the detection of flares may be dependent on the duration of follow-up). There was no difference in the frequencies of anti-Ro/La antibody (89% compared with 72%) or rheumatoid factor (RF) positivity (78% compared with 72%), whereas the flare group were slightly more likely to be antinuclear antibody (ANA) positive [8/9 (89%)] compared with those without [26/50 (52%); P = 0.039]. The proportion of patients in this cohort with fibromyalgia was under 5% [4]. Table 1 summarizes the clinical features of the flares. These were predominantly musculoskeletal and consisted of: increased fatigue/malaise, 7/12 (58%); polyarthritis, 7/12 (58%); polyarthralgia, 7/12 (58%); myalgia, 2/12 (17%); alopecia, 1/12 (8%); tenosynovitis, 1/12 (8%) and monoarthritis, 1/12 (8%). In addition, two patients reported worsening Raynaud's phenomenon [2/12 (17%)] and one worsening livedo reticularis [1/12 (8%)]. In two patients with known osteoarthritis, the features of the flare could not obviously be accounted for by this nor by an exacerbation of fibromyalgia. Clinical details of ‘flares’ in nine female patients with primary Sjögren's syndrome Clinical details of ‘flares’ in nine female patients with primary Sjögren's syndrome No clear pattern of changes were observed for the erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), immunglobulin (Ig) G, IgA and IgM levels, or complement C3 or C4 levels in accord with previous studies [5]. Data on anti-Ro and anti-La antibody titres were available in five patients and, similarly, no clear pattern was seen. Eight flares settled following intramuscular depomedrone injections. In six patients, hydroxychloroquine therapy was initiated and one patient was started on oral prednisolone. Non-steroidal anti-inflammatory drugs (NSAIDs) were prescribed for two patients. All flares settled within 9 months. The overall frequency of flares in our cohort at 3–7% per year is much less than that in SLE (62% per year [2]) and the frequency of flares involving the general or musculoskeletal systems also appears substantially lower than in SLE [2]. We did not attempt to assess the severity of the flares in relation to the need for therapy but subjectively they appeared modest in severity and settled over weeks to months with modest therapy. In particular, we did not see any exacerbations involving internal organ systems such as pulmonary or haematological features, although these have been well described in other cohorts [6]. Although the numbers were too small for statistical analysis, two out of three flares that were not treated lasted for over 6 months (Table 1), compared with one out of nine for which treatment was given. This descriptive study is limited by its retrospective nature and lack of data from comparison groups. It does, however, provide preliminary data to guide the design of standardized activity and damage measures for PSS based on the BILAG and SLICC measures used in SLE [7, 8]. These should enable us to characterize the systemic features of PSS and their variation over time in more detail in a prospective study. The authors are not aware of any conflicts of interest.