AB0702 TREATMENT EFFECT OF TUMOR NECROSIS FACTOR A INHIBITORS ON MAGNETIC RESONANCE IMAGING PROGRESSION IN PATIENTS WITH SPONDYLOARTHRITIS: A META-ANALYSIS

Abstract

Background Treatment effect of tumor necrosis factor α inhibitors (TNFi) in patients with spondyloarthritis (SpA) had been proved by plenty of studies that could remarkably eliminate symptoms, signs, and laboratory inflammatory manifestations [1,2]. Simultaneously, the treatment effect of TNFi on delaying radiographic progression in SpA was often reported [3,4], including magnetic resonance imaging (MRI) which has a high sensitivity on detecting the earliest inflammation [5]. Objectives A meta-analysis was performed to summarize the treatment effect of TNFi on MRI progression reflected by SPARCC score in SpA patients Methods Comprehensive search was conducted in the electronic databases of OVID Medline, OVID EMBASE and Cochrane library on Nov.27, 2018. All randomized controlled trials (RCTs) focused on MRI progression and disease activity in SpA patients treated with TNFi were included. Primary outcome was the Spondyloarthritis Research Consortium Canad (SPARCC) MRI scoring system, accompanied with or without other outcomes, including ankylosing Spondylitis Disease activity Score (ASDAS), Bath ankylosing Spondylitis disease activity index (BASDAI), Bath ankylosing Spondylitis functional index (BASFI), C-reactive protein (CRP). Data were pooled by mean differences (MD) with 95% confidence intervals (CI) and publication bias was assessed by funnel plot. Sensitivity analysis was performed to test the result robustness. Jadad scale was applied to assess the methodology quality of included trials. Two reviewers independently selected studies, extracted data, and assessed study quality. Results Totally 10 RCTs enrolled 1006 patients with high methodology were included. Compared with control group, TNFi significantly improved SPARCC score at sacroiliac joint (SIJ) (MD=2.84, 95% CI 2.43-3.25), SPARCC score of spine (MD=1.87, 95% CI 1.27-2.46), aSDAS (MD=0.96, 95% CI 0.71-1.20), BASDAI (MD=1.15, 95% CI 0.51-1.78), BASFI (MD=0.95, 95% CI 0.51-1.40), and CRP (MD=4.64, 95% CI 1.06- 8.23) in double-blind phase. Subgroup analyses by disease subgroup and individual TNFi showed treatment effect of TNFi on delaying MRI progression was not affected by disease subgroup and individual TNFi. Sensitivity analysis showed the treatment effects of TNFi versus placebo were consistent with TNFi versus controls, suggesting the results of the study were robust. The funnel plot of SPARCC score at SIJ based on TNFi versus control in double-blind phase was asymmetric, suggesting there might have potential publication bias. Conclusion TNFi are effective to treat SpA patients and delay MRI progression which are assessed by SPARCC score and treatment effect is consistent among disease subgroups and individual TNFi.