Non‐genetic biomarkers for ankylosing spondylitis: An umbrella review

OBJECTIVE: The objective of the study was to provide an overview of the existing evidence on non‐genetic biomarkers for ankylosing spondylitis (AS). METHODS: In this umbrella review, we searched PubMed and Web of Science from database inception to October 31, 2020. Systematic reviews and meta‐analyses of observational studies investigating the associations between any non‐genetic biomarkers and AS were included. We estimated summary standardized mean difference (SMD) along with 95% confidence interval (CI), I ( 2 ) statistic, 95% prediction interval (PI), and assessed small‐study effects and excess significance bias. The study was registered in PROSPERO with registration number of CRD42020218240. RESULTS: A total of 1276 publications were identified, of which 21 articles covering 43 non‐genetic biomarkers were eligible for inclusion. Evidence of 22 (51%) non‐genetic biomarkers exhibited a nominally significant effect (p < 0.05) on AS, and 7 associations (14%) showed small‐study effects. The associations of platelet count (SMD: 0.53, 95% CI: 0.36 to 0.71) and serum interleukin (IL)‐23 levels (SMD = 2.03, 95% CI: 1.27 to 2.79) with AS presented highly suggestive evidence, while circulating IL‐17 levels (SMD = 2.36, 95% CI: 1.71, 3.01) and Treg/PBMC ratio (SMD = −0.75, 95% CI: −1.06 to −0.44) presented suggestive evidence. However, these associations showed large or very large between‐study heterogeneity, suggesting an indefinite direction for the effect when 95% PIs were considered. CONCLUSION: No convincing evidence supported the existence of a non‐genetic biomarker for AS. Some highly suggestive associations might be affected by bias, therefore, promising non‐genetic biomarkers for AS remain limited at least based on the current evidence from observational studies.