Relationship between glucocorticoid dose and adverse events in systemic lupus erythematosus: data from a randomized clinical trial

Abstract

Objective: The aim was to obtain a better understanding of the relationship between exposure to glucocorticoids (GCs) and adverse events (AEs) reported in a randomized controlled trial (RCT) in systemic lupus erythematosus (SLE) patients.

Method: We used data from the BLISS-76 trial on belimumab. The data were accessed through an agreement under the SHARE mechanism. AEs were grouped according to medical relevance, i.e. based on similarity of symptoms and pathophysiology. We studied the relationship between AEs and exposure to GCs at baseline and at any time-point, and compared the frequencies of each AE and groups of AEs between tertiles of cumulative GC dose.

Results: In total, 991 AEs were reported in the 819 patients of the trial. The frequencies of anaemia, pyrexia, oral herpes, and malaise were significantly higher (p < 0.05) in patients who were on GCs at baseline than in those who were not. The frequencies of several other AEs, including nausea, seasonal allergy, bacterial sinusitis, and viral upper respiratory infection, were significantly higher in patients without GCs. For cumulative GCs, tachycardia and proteinuria were significantly more frequent in the highest tertile than in the lowest tertile, but other AEs and groups of AEs were significantly more frequent in the lowest tertile.

Conclusion: This study highlights the feasibility of post-hoc analyses of RCTs using the SHARE mechanism and demonstrates the association of GCs with various AEs. Contrary to expectations, there were also associations between lower cumulative GC dose and several other AEs.

Acknowledgements

We thank Dr A Thompson at GSK for helping us with implementation of the SHARE collaboration, and Dr M Backheden at Karolinska Institutet for his expert statistical advice.