Diagnostic accuracy of enzyme-linked immunosorbent assays (ELISA) to detect anti-skin autoantibodies in autoimmune blistering skin diseases: a systematic review and meta-analysis.

Background: Systematic reviews and meta-analysis are essential tools to accurately and reliably summarize evidence, and can be used as a starting point for developing practice guidelines for the diagnosis and treatment of patients.

Aim: To estimate the diagnostic accuracy of enzyme-linked immunosorbent assays (ELISA) to detect anti-BP180 and anti-desmoglein 3 (Dsg3) autoantibodies in the diagnosis of autoimmune blistering skin diseases.

Methods: A Medline search of English written articles, published between 1994 and 2011, reporting data on the sensitivity and specificity of diagnostic tests was conducted using the following search terms: "BP180 autoantibodies", "Dsg3 autoantibodies", and "enzyme linked immunosorbent assay". The selected articles have been evaluated according to the quality of the statistical methods used to calculate diagnostic accuracy (definition of cutoff value, use of ROC curves, and selection of control cases). The meta-analysis was performed using a summary ROC (SROC) curve and a random-effect model to independently combine sensitivity and specificity across studies.

Results: The search yielded 69 publications on BP180 autoantibodies and 178 on Dsg3 autoantibodies. A total of 30 studies met the inclusion criteria: 17 provided data on the assays to detect autoantibodies to BP180 in a sample of 583 patients with bullous pemphigoid (BP), while 13 studies provided data on the assays to search for anti-Dsg3 autoantibodies in a sample of 1058 patients with pemphigus vulgaris (PV). The 17 studies on BP180 autoantibodies yielded a pooled sensitivity of 0.87 (95% confidence interval (CI) 0.85 to 0.89) and a pooled specificity of 0.98 (CI, 0.98 to 0.99). The area under the curve (AUC) for the SROC curve was 0.988, and the summary diagnostic odds ratio was 374.91 (CI, 249.97 to 562.30). The 13 studies on Dsg3 autoantibodies which met the inclusion criteria, yielded a pooled sensitivity of 0.97 (CI, 0.95 to 0.98), and a pooled specificity of 0.98 (CI, 0.98 to 0.99). The AUC for the SROC curve was 0.995 and the summary diagnostic odds ratio was 1466.11 (95% CI, 750.36 to 2864.61).

Conclusions: Results of the meta-analysis demonstrated that ELISA tests for anti-BP180 and anti-Dsg3 autoantibodies have high sensitivity and specificity for BP and PV, respectively, and can be used in daily laboratory practice for the initial diagnosis of autoimmune blistering skin diseases.