α-Synuclein seed amplification assay methodology and performance in Parkinson's disease, lewy body dementia, and...

Seed amplification assays have shown promise in research for accurate diagnosis of synucleinopathies. In consideration of clinical implementation, gaps in the literature include that performance data are frequently determined using clinically unrelated controls (e.g., healthy controls or phenotypically unrelated conditions [UC]), and a lack of emphasis on methodological variability, including required replicates and positivity thresholds. A review and meta-analysis were performed to assess the methodological parameters and diagnostic performance of seed amplification assays for detecting synucleinopathies and, where necessary, cohorts were adjusted to be more representative of the populations in which the testing would be deployed in clinical practice. A search was conducted for α-synuclein seed amplification assay studies on Parkinson's disease, dementia with Lewy bodies, and multiple system atrophy, for matrices including cerebrospinal fluid (CSF), skin, and olfactory mucosa (OM). Assay methodological details were extracted, as were diagnostic performance data. For the latter, negative controls were divided into two distinct groups: disease mimics (DM) and UC. A total of 55 studies met the inclusion/exclusion criteria. Methodological parameters varied including the concentration, sequence and source of the assay substrate, as well as required assay replicates and determination of the positivity threshold. Median sensitivities and specificities relative to DM groups for CSF were 0.92 (95% confidence interval: 0.88-0.96) and 0.90 (0.89-0.96), for skin were 0.94 (0.79-1.0) and 0.86 (0.83-1.0), and for OM were 0.69 (0.33-1.0) and 0.94 (0.83-1.0), respectively. Although diagnostic performance was slightly reduced when adjusting for clinically relevant populations, it remained encouragingly high. Towards broader clinical implementation, valuable research directions include further streamlining of analytical workflows, and characterizing diagnostic performance by stage of disease and co-pathologies.