Diagnostic accuracy of anti-integrin αvβ6 in ulcerative colitis: a diagnostic meta-analysis

Ulcerative colitis (UC) presents with diagnostic challenges due to symptom overlap with other gastrointestinal (GI) disorders. Recent studies identify anti-integrin αvβ6 autoantibodies as a promising biomarker for UC. This meta-analysis aims to evaluate the diagnostic accuracy of anti-integrin αvβ6 antibodies in distinguishing UC from healthy individuals and other GI diseases.

We conducted a systematic literature search of PubMed, Scopus, and Embase up to February, 2025, following PRISMA guidelines. Studies assessing the diagnostic performance of anti-integrin αvβ6 antibodies in UC patients were included. A bivariate random-effects model was used to pool sensitivity and specificity estimates using STATA. Forest plots and SROC curves were generated. Meta-regression and interaction analyses explored the influence of covariates such as control group type, age group, and geographic region on diagnostic performance. Risk of bias was assessed using tool QUADAS-2. Post hoc analyses were conducted to assess the impact of cut-off thresholds and ELISA platforms on the diagnostic performance of anti-integrin αvβ6 antibodies.

Six studies comprising 3887 participants (1904 UC patients) were included in meta-analysis. The pooled sensitivity and specificity of anti-integrin αvβ6 for UC across all comparator groups were 83% (95% CI: 0.70-0.91) and 93% (95% CI: 0.88-0.97), respectively. Diagnostic performance remained consistent across control types for sensitivity but varied significantly for specificity, especially when Crohn's disease was used as a comparator (81%; 95% CI: 0.75-0.86). Multivariate meta-regression identified patient age, geographic region, and control group type as significant modifiers of specificity. Interaction models further confirmed a combined influence of these factors on diagnostic performance. Post hoc analyses revealed that sensitivity remained stable across thresholds (2SD: 0.65, 3SD: 0.87), while specificity varied significantly depending on cut-off values (2SD: 0.89, 3SD: 0.92) and ELISA methodology (0.92 vs. 0.83).

Our meta-analysis demonstrates that anti-integrin αvβ6 antibodies exhibit high diagnostic accuracy for UC, with consistent sensitivity and specificity. Their performance is influenced by patient demographics and study region, suggesting the need for tailored diagnostic criteria in clinical settings.