In lupus, infection risk tracks the steroid dose more than the immunosuppressant you choose

The edition in brief

A target trial emulation in 6,168 US patients with non-renal lupus starting azathioprine, belimumab, methotrexate or mycophenolate found two-year infection-related hospitalisation of 8–13%. Mycophenolate looked riskier than belimumab (HR 1.55) and methotrexate at first, but the differences disappeared once time-varying glucocorticoid dose was accounted for — steroid exposure, not the steroid-sparing drug, carried the risk. The FDA has listed a Class I recall of Tyenne (tocilizumab-aazg) 400 mg/20 mL vials made by Fresenius Kabi in the US after glass particles were found; it is a manufacturing defect, not a safety signal for tocilizumab. In 70 Chinese patients with lupus-associated warm autoimmune haemolytic anaemia, B-cell-targeted therapy and conventional treatment gave similar six-month response (89.7% vs 85.7%), with lower steroid doses at one month. A Norwegian register of 2,932 adults with juvenile idiopathic arthritis found more hypertension, type 1 diabetes, coeliac and thyroid disease and higher mortality than matched controls. A meta-analysis of 4,106 children with Kawasaki disease found adjunctive corticosteroids did not significantly reduce coronary abnormalities, on very-low-certainty evidence.