TY - JOUR T1 - Risk of adverse events from different drugs for SLE: a systematic review and network meta-analysis JF - Lupus Science & Medicine DO - 10.1136/lupus-2017-000253 VL - 5 IS - 1 SP - e000253 AU - Jingru Tian AU - Yien Luo AU - Haijing Wu AU - Hai Long AU - Ming Zhao AU - Qianjin Lu Y1 - 2018/03/01 UR - http://lupus.bmj.com/content/5/1/e000253.abstract N2 - Objective The comparative safety of immunosuppressive drugs, biologicals and glucocorticoids (GC) for patients with SLE remains controversial. We aimed to investigate the specific side effects of the available SLE drugs in this population of patients.Methods Electronic databases were systematically searched through September 2017 for randomised trials in patients with SLE. The primary outcomes were all-cause mortality and withdrawal related to adverse events (AEs). We performed a random-effects network meta-analysis to obtain estimates for primary and secondary outcomes and presented these estimates as ORs with 95% CIs.Results Forty-four studies comprising 9898 participants were included in the network meta-analysis. No drug regimen was considered to be safer for reducing all-cause mortality. However, compared with cyclophosphamide, azathioprine (OR 3.04, 95% CI (1.44 to 6.42)) and cyclosporine (OR 3.28, 95% CI (1.04 to 10.35)) were significantly less safety in AE-related withdrawals, and GC was ranked lowest and led to higher withdrawal rates. Tacrolimus (TAC) was ranked high and showed a benefit in many outcomes. Biologicals and chloroquine also showed good safety in all of the available outcomes, while the beneficial effects of other immunosuppressive drugs were not substantial in different types of serious adverse events.Conclusions TAC is the safest strategy for patients with SLE. Biologicals and chloroquine are also fairly safe for patients with SLE. The use of other immunosuppressive drugs and GC needs to be balanced against the potential harms of different types of AEs, and the practical safety of drug combinations still requires further trials to evaluate. ER -