Amyloid Plaques in Retina for Diagnosis in Alzheimer's Patients: a Meta-Analysis.

Background: Detection of retinal β-amyloid (Aβ) peptide accumulation is a novel diagnostic method for Alzheimer's disease (AD), but there is, as yet, no conclusive evidence of its accuracy. Aim: To identify the diagnostic accuracy of pathological retinal Aβ detection for AD by a meta-analytic approach. Methods: Electronic and reference searches were conducted to identify studies related to the diagnostic effects of retinal Aβ detection in AD that met pre-defined inclusion criteria. The QUADAS-2 tool was employed to assess the risk of bias, and Review Manager plus the Open Meta-Analyst were used to perform the data analysis. Results: From 493 unduplicated reports, five studies with small sample sizes were included in this review. Six staining methods were employed. The eligible studies showed extremely broad ranges of sensitivity (0-1.00) and specificity (0.50-1.00) with substantial heterogeneity. The estimates of positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR) were also extremely varied (from 0.71 to 11.57 for PLR, from 0.04 to 1.11 for NLR, and from 0.69 to 297.00 for DOR). Conclusions: The limited number of eligible studies and their methodological heterogeneity make it impossible to come to a conclusion whether pathological retinal Aβ detection is an effective diagnostic tool for AD. More studies, especially large surveys investigating retina Aβ load with quantitative methods among consecutive or random samples, are needed to determine the accuracy of Aβ detection for diagnosing AD.