Rituximab in Sjögren’s disease: reconciling evidence from randomized and observational studies

Sjögren’s disease is a chronic, systemic autoimmune disease characterized by the lymphocytic infiltration of exocrine glands, primarily salivary and lacrimal glands, leading to sicca symptoms and various extraglandular manifestations [1,2]. Despite improved understanding of its immunopathogenesis, effective disease-modifying therapies for Sjögren’s disease remain elusive. Given the central role of B cells in disease initiation and propagation, evident through B-cell hyperactivity, germinal center-like structures in glands, and increased risk of B-cell lymphoma, targeting B cells has emerged as a rational therapeutic approach [3]. Thus, rituximab (RTX), an anti-CD20 monoclonal antibody that selectively depletes B cells, has garnered considerable attention. However, clinical trials evaluating RTX in Sjögren’s disease have yielded mixed results, prompting further investigations into the true magnitude and consistency of its therapeutic effect [4,5].

A recent systematic review and meta-analysis synthesized evidence from both randomized controlled trials (RCTs) and quasi-experimental studies (i.e., before-after trials without comparator arms) to clarify RTX’s efficacy and safety in Sjögren’s disease [6]. This combined approach reflects a pragmatic strategy to overcome the scarcity of RCTs in rare autoimmune diseases while maintaining critical evaluations of study quality and heterogeneity. A prominent finding of this meta-analysis was the discordance in observed efficacy between RCTs and quasi-experimental studies. Among the 15 included studies (4 RCTs, 11 quasi-experimental), pooled results from quasi-experimental studies demonstrated a significant reduction in disease activity after RTX treatment, as measured by the European League Against Rheumatism Sjögren’s Syndrome Disease Activity Index (ESSDAI). Specifically, the mean difference in ESSDAI from baseline to post-treatment was –4.36 (95% confidence interval [CI]: –5.83 to –2.89), indicating meaningful clinical improvement. In contrast, the pooled analysis of RCTs failed to demonstrate a statistically significant difference in ESSDAI between RTX and placebo groups (mean difference: 0.09; 95% CI: –0.43 to 0.61), suggesting that RTX may not confer substantial benefit over placebo in rigorously controlled settings. Changes in ESSDAI scores reflect fluctuations in systemic disease activity in Sjögren’s disease, and accumulating evidence suggests that persistent or recurrent high activity is associated with poorer long-term outcomes. Patients with inadequate ESSDAI improvement are more likely to develop irreversible organ damage, particularly affecting pulmonary, neurological, or renal systems. Moreover, insufficient disease control correlates with worse patient-reported outcomes, including fatigue, pain, and impaired quality of life. Thus, monitoring ESSDAI trajectories provides prognostic insight beyond short-term disease activity, linking treatment response with long-term morbidity.

This discrepancy raises critical questions regarding trial design, patient selection, and the interpretation of treatment effects in autoimmune disease research. While RCTs remain the gold standard for evaluating therapeutic efficacy due to their ability to minimize bias through randomization and blinding, they are often limited by strict inclusion criteria, relatively short follow-up durations, and underpowered sample sizes [7]. The patients in the reviewed RCTs had a longer mean disease duration (93.6 months) than those in quasi-experimental studies (59.9 months), possibly indicating more established, fibrotic disease less amenable to immunomodulation. Conversely, quasi-experimental studies, often conducted in real-world settings, may include a broader range of disease phenotypes and treatment contexts, but are inherently more susceptible to biases, particularly selection bias, observer bias, and regression to the mean. Importantly, the open-label design of quasi-experimental studies allows for the potential overestimation of treatment effects due to expectation bias and the lack of blinded outcome assessments [8]. Given the nature of current classification criteria, recruiting early-stage patients is inherently difficult, which likely contributes to the long disease duration observed in both RCTs and other included studies [2]. Publication bias is also a plausible concern, as positive results from uncontrolled trials are more likely to be published than negative or neutral findings. These methodological limitations necessitate the cautious interpretation of the apparent superiority of RTX in quasi-experimental studies.

Marked heterogeneity was observed across studies, particularly within RCTs. This heterogeneity likely reflects diverse clinical settings, RTX dosing schedules, concomitant immunosuppressive use, and variable disease manifestations. The notion that RTX may benefit select subgroups of Sjögren’s disease patients remains plausible. Anecdotal and subgroup data suggest greater benefit in patients with hypergammaglobulinemia, rheumatoid factor positivity, or early systemic involvement. However, such observations remain hypothesis-generating, and future trials should incorporate stratification by immunological or clinical biomarkers to elucidate potential predictors of response. The current evidence suggests that RTX may offer meaningful benefit for certain Sjögren’s disease patients, particularly in real-world settings, but its efficacy in a broader, unselected population remains uncertain. These findings underscore the need for personalized approaches in Sjögren’s disease management, incorporating disease characteristics, patient preferences, and potential biomarkers of therapeutic response. Given the conflicting outcomes between RCTs and quasi-experimental studies, future research must prioritize the design of adequately powered, multicenter RCTs with stratified randomization and adaptive elements. Pragmatic trials that better reflect clinical practice, while maintaining methodological rigor, could help reconcile real-world observations with trial data.

In conclusion, RTX remains a biologically rational and potentially beneficial therapy for Sjögren’s disease, yet robust evidence supporting its widespread use is currently lacking. While quasi-experimental studies report favorable outcomes, results from high-quality RCTs have been inconclusive, reflecting a complex interplay of methodological, biological, and clinical factors. The findings of this meta-analysis highlight the need for more rigorous and targeted research to identify which patients are most likely to benefit from RTX, the optimal timing and duration of therapy, and strategies to enhance both efficacy and safety. Until then, the use of RTX in Sjögren’s disease should be guided by careful clinical judgment and individualized assessments.