High Agreement Across Laboratories Between Different Alpha-Synuclein Seed Amplification Protocols

Abstract

Background: Seed amplification assays (SAA) detect alpha-synuclein (aSYN) pathology in patient biomatrices such as cerebrospinal fluid (CSF)-potentially even before clinical manifestations. As CSF-based SAA are approaching broader use in clinical trials and research, ensuring that different laboratories obtain the same results becomes increasingly important.

Methods: In this cross-laboratory, cross-aSYN-recombinant substrate and cross-protocol round-robin test, we compared SAA results from a common set of 38 CSF samples measured independently in four research laboratories of the German Center for Neurodegenerative diseases. Three laboratories (A-C) used an assay protocol adapted from Parchi's group at ISNB (Bologna, Italy); laboratory D used an assay protocol adapted from Amprion Inc. Two different manufacturers of aSYN protein were used as substrates for the SAA reaction.

Results: Qualitative results were identical in at least three of the four laboratories for 37 out of 38 samples (20 positive, 17 negative). Fleiss Kappa for all four laboratories was 0.751 (z = 12, p < 0.001). For each laboratory, agreement with laboratory A was > 92%. For the number of positive replicates, Fleiss Kappa was 0.45 for a score of zero positive replicates and 0.42 for a score of four positive replicates.

Conclusions: The qualitative SAA results showed a high level of agreement across research laboratories, aSYN monomers, and assay protocols. Small differences between laboratories were systematic, consistent with the notion that SAA reports biologically relevant properties. These results also underline that round-robin tests can be helpful in assessing and ensuring SAA quality across laboratories.

Keywords: RT‐QuIC; alpha‐synuclein; method validation; proficiency testing; ring trial.

FIGURE 1

(a) Qualitative results from SAA measurements of 38 samples in 4 different laboratories. Results from laboratories A–C were either “Non‐Seeder” or “Seeder”; results from laboratory D were “Non‐Seeder” or “(Type 1) Seeder” or “Type 2 Seeder”. Two measurements of distinct samples were inconclusive by one laboratory (gray). (b) Quantitative results. In laboratories A–C, each sample was run as quadruplicates. The number of positive replicates for each sample is color‐coded.

FIGURE 2

Correlation of kinetic parameters between laboratories. (a) Correlation of the second fastest lag phase (TT2) measured in laboratory A with TT2 measured in laboratory B. (b) Correlation of average lag phase (LAG) measured in laboratory A with LAG measured in laboratory B. (c) For samples positive in one laboratory and negative in the other, TT2 was set to 40 h (TT2neg40). These modified values for TT2 were then correlated between laboratories A and B. In panels (a–c), each marker represents one patient sample.