Efficacy and safety of biological agents in the treatment of patients with Takayasu arteritis: a systematic review and meta-analysis.

OBJECTIVE: We aimed to systematically review biological agents' efficacy and safety in patients with Takayasu arteritis (TAK). MATERIALS AND METHODS: A systematic literature search of 7 electronic databases, including MEDLINE (via PubMed), EMBASE, Elsevier ScienceDirect, EBSCO, Springer Link, Web of Science, and Cochrane Library on the efficacy of biological agents on patients with TAK was conducted. Only studies published in English and with a sample size >5 patients with TAK were included. Two reviewers independently selected studies, extracted data and assessed its methodological quality. Random effects meta-analyses of various effect measures were performed. RESULTS: According to the title and abstract, 961 studies were identified and screened. Subsequently, 31 studies from 29 observational studies and 2 randomized-controlled trials (RCTs), which included a total of 517 patients with TAK that met the inclusion and exclusion criteria, were selected. Observational studies showed a high risk of bias. Pooled remission rates of biological agents were 66% (95% CI: 58%-73%; I2=59%), and the remission rates of anti-tumor necrosis factor (TNF) agents and tocilizumab (TCZ) were similar: 65% (95% CI: 56%-73%; I2=49%) and 70% (95% CI: 55%-86%; I2=69%), respectively. Pooled relapse rates were 23% (95% CI: 15%-31%; I2=66%). The relapse rate was 28% (95% CI: 16%-40%; I2=68%) for anti-TNF agents and 17% (95% CI: 7%-26%; I2=49%) for TCZ. The remission rate of TCZ was slightly higher (p>0.05), but the relapse rate was statistically significantly lower than that of anti-TNF agents (p=0.017). Furthermore, biological agents significantly decreased the doses of glucocorticoid (GC) and levels of acute phase inflammation markers (ESR, CRP) while the proportion of patients with new angiographic lesions or progression of previously noted lesions were 11% (95% CI: 4%-18%; I2=59%). RCTs with a small sample size showed abatacept was ineffective, and TCZ was underpowered to detect a difference in time to relapse compared to placebo. The most common adverse event of biological agents was infection (6%, 95%CI: 2%-10%). No deaths were reported. CONCLUSIONS: Although the beneficial effects of biological agents are encouraging in enhancing disease remission, reducing the levels of acute phase inflammation markers and decreasing the treatment doses of GC in patients with TAK, there is still a risk of relapse. More refined studies with larger cohorts are necessary before drawing a definitive opinion.