Assays and protocols

How well does CSF EP-QuIC perform in routine use?

Simon and colleagues (Canadian Journal of Neurological Sciences, 2021) studied 716 prospectively collected Canadian CSF samples submitted between April 2017 and June 2019. After exclusions for insufficient volume, xanthochromic or bloody samples, pending diagnoses and indeterminate QuIC results, 623 remained: 98 autopsy-confirmed definite CJD, 24 probable CJD, 222 with alternative diagnoses and 279 classified as unknown but not suspected CJD. Positive predictive values were 96 percent for EP-QuIC, 68 percent for the 14-3-3 gamma immunoassay and 66 percent for the total tau immunoassay. The authors state that the assays are accredited under ISO/IEC 17025 and offered free of charge to Canadian health care providers, with samples sent to the Prion Diseases Section of the National Microbiology Laboratory. This is a single national laboratory, and the study reports positive predictive value for its own sample set.

Can a modified assay detect inherited and rarer forms?

Myskiw and colleagues (Annals of Neurology, 2026) report a retrospective comparison on CSF from 61 sporadic CJD, 50 inherited prion disease and 5 variably protease-sensitive prionopathy cases, with 71 pathologically confirmed non-CJD controls used in the initial substrate comparison. Their modified assay, which they call ES-QuIC, replaces the hamster substrate with a recombinant deer mouse protein.

By subgroup, ES-QuIC detected all 61 sporadic cases, 12 of 13 fatal familial insomnia cases (standard EP-QuIC 4 of 13, IQ-CSF none), all 12 Gerstmann-Straussler-Scheinker cases (IQ-CSF 4 of 12), 5 of 5 VV1 and 5 of 5 MM2 cases, and 3 of 5 VPSPr cases with 1 indeterminate. The one V180I case was detected by IQ-CSF and missed by both end-point assays.

The authors state these limits: a retrospective design, one Canadian diagnostic laboratory with a US surveillance centre supplying samples, subgroups of 1 to 5 cases, no serial samples, limited specificity data for the new substrate, and no independent or inter-laboratory validation yet. Some comparisons are reported as data not shown.

What about blood?

Thomas and colleagues (Blood, 2024) compared RT-QuIC and two versions of protein misfolding cyclic amplification on longitudinal blood series from a large sheep model of variant CJD, in a blinded design. Both cyclic amplification versions had 100 percent sensitivity and RT-QuIC 61 percent. All three detected infection in samples taken 26 months before clinical signs, with no false positives reported. The authors suggest the assays may suit detection of asymptomatic vCJD infection in humans. These are sheep samples, and human blood performance is not shown by this study.

What is not covered?

Assays for decontamination, tissue and food-chain testing, and national reference methods were not read for this page. Lists of reference laboratories are in the Atlas, for example the Ottawa prion reference laboratory, ECDC, the EU reference laboratory in Turin, Hospital Clinic Barcelona and ISS Rome. Related pages are antibodies and reagents, biobanks and the resources hub.