Efficacy and safety of golimumab in ankylosing spondylitis and axial spondyloarthritis: a systematic review and meta-analysis of randomized controlled trials

Abstract

Background Golimumab is an approved tumor necrosis factor inhibitor for the treatment of ankylosing spondylitis (AS) and axial spondyloarthritis (axSpA). Although randomized controlled trials have demonstrated its efficacy, evidence regarding long-term treatment maintenance remains limited. This systematic review and meta-analysis evaluated the efficacy, safety, and durability of golimumab therapy in patients with AS and axSpA. Methods This review was conducted according to PRISMA 2020 guidelines and prospectively registered in PROSPERO (CRD420251130466). PubMed, ScienceDirect, Scopus, and Google Scholar were searched from inception to January 2026. Randomized placebo-controlled trials of subcutaneous or intravenous golimumab were pooled using random-effects risk ratios (RRs) with Hartung–Knapp 95% confidence intervals (CIs), analyzing the approved dose in multi-arm trials; extension studies were synthesized separately as single-arm proportions. Primary outcomes were ASAS20, ASAS40, BASDAI50, and serious adverse events (SAEs). Risk of bias was assessed with RoB 2 (trials) and the NHLBI before–after tool (extensions), and certainty with GRADE. Results Seven publications representing four independent trials (834 randomized patients) were included. At Weeks 14–16, golimumab increased ASAS20 (RR 2.22, 95% CI 1.51–3.27; high certainty), ASAS40 (RR 3.25, 95% CI 1.89–5.59; high certainty), and BASDAI50 responses (RR 2.82, 95% CI 1.39–5.71; moderate certainty), corresponding to approximately 300–350 additional responders per 1,000 patients treated. Effects favored golimumab in both radiographic AS and non-radiographic axSpA, with no statistically significant subgroup difference (ASAS20, p = 0.065). Evidence on SAEs was very imprecise (three trials, 14 events; RR 0.83, 95% CI 0.08–8.54; low certainty), and the risk of any adverse event was similar to placebo (RR 1.05, 95% CI 0.85–1.29). In uncontrolled extensions, 72% (95% CI 35%–92%) and 63% (95% CI 31%–86%) of patients originally randomized to golimumab had ASAS20 and ASAS40 responses at Weeks 52–104 (very low certainty). Conclusion High- to moderate-certainty randomized evidence shows that golimumab substantially improves short-term signs and symptoms of AS and axSpA compared with placebo. Safety data are too sparse to exclude important differences in serious adverse events, and evidence on durability of response is of very low certainty. All included trials were industry-sponsored, and independent long-term comparative studies are needed. Systematic Review Registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420251130466 , PROSPERO CRD420251130466.