Highlights
-
• EP-QuIC testing remains accurate even when testing rates are increased significantly.
-
• Expanded regional testing improved surveillance and revealed a modest increase in VV2 CJD cases.
-
• Improving access to EP-QuIC testing enhances prion disease detection.
Introduction
Sporadic Creutzfeldt-Jakob disease (sCJD) is an invariably fatal prion disease caused by spontaneous misfolding of the normally healthy prion protein (PrP) resulting in an accumulation of detergent-insoluble protein structures in the brain and associated neuropathological changes. Historically, non-specific biomarkers including 14-3-3γ and total tau, which are neuronal proteins released during degeneration, have been relied on in the clinic to assign a diagnosis of possible or probable CJD in living patients, along with specific clinical features, neuroimaging and radiographic findings.Reference Zerr, Kallenberg and Summers1 Recently, the development of the highly sensitive and specific quaking-induced conversion (QuIC) assay has transformed the antemortem diagnosis of CJD by enabling direct detection of infectious prion seeds in cerebrospinal fluid (CSF).Reference Orrú, Groveman, Hughson, Zanusso, Coulthart and Caughey2,Reference Atarashi, Satoh and Sano3 The presence of a progressive neuropsychiatric disorder and a positive QuIC test is now considered indicative of probable CJD.Reference Hermann, Laux and Glatzel4–Reference Rhoads, Wrona and Foutz6 In Canada, CSF testing for CJD is centralized federally within the Public Health Agency of Canada, with ISO-accredited testing performed exclusively at the National Microbiology Laboratory. Sample acceptance is not restricted on the basis of any clinical criteria, since limited or no clinical information is provided at the time of testing. The CSF submission processes however can still vary provincially depending on local biosafety risk assessments, more- or less-strict adherence to case definitions prior to test requisition, and varying levels of dependence on centralized provincial labs to screen and route samples for analysis. All of these factors can influence the demographics associated with submitted samples across provinces. Once submitted, CSF specimens are tested by end-point QuIC (EP-QuIC),Reference Cheng, Vendramelli and Sloan7–Reference Simon, Peterson and Phillipson9 and until 2025, in parallel for levels of 14-3-3γ and total tau.Reference Slota, Myskiw and Peterson10
CJD is a highly heterogeneous condition which can make it challenging to diagnose. While the classic presentation of rapidly progressive dementia, myoclonus and cerebellar ataxia, combined with characteristic MRI, are strongly suggestive of CJD, atypical presentations can mimic more common neurodegenerative syndromes.Reference Hermann, Appleby and Brandel5 Classic CJD is apparent in approximately 80% of cases, whereas atypical presentations characteristic of some sporadic molecular subtypes and genetic cases make up the remainder. Earlier diagnosis is important to direct treatments, manage symptoms and to arrange end-of-life care once CJD is identified. The introduction of the QuIC test, which has proven to be extremely specific for misfolded PrP, provides an avenue for routine clinical testing during differential diagnosis of neurological syndromes of uncertain etiology. While QuIC has been used for a number of years in several specialized CJD testing laboratories,Reference Rhoads, Wrona and Foutz6,Reference Slota, Myskiw and Peterson10–Reference Hermann, Schmitz and Cramm13 it is generally applied to cases for confirmatory testing once the clinical criteria for classic CJD have been met. The use of the QuIC test as a more general tool in differential diagnosis has not been thoroughly explored. Caution in its use would be the identification of “false positive” tests, given that a diagnosis of CJD is a fatal and untreatable prognosis.
We recently evaluated the performance of EP-QuIC against a cohort of definitively diagnosed CJD cases confirmed by post-mortem neuropathological evaluation.Reference Slota, Myskiw and Peterson10 During this evaluation, we reported a high volume of CSF submissions – over five times the national per-capita average – originating from the province of New Brunswick (NB). We attributed this to a local increase in prion disease awareness driven by reports of a neurological syndrome of unknown cause (NSUC) in the region,14 originally postulated to be a prion disease. Recently, an independent evaluation of 25 cases from the cohort initially diagnosed with NSUC failed to support the existence of a specific novel neurological disorder.Reference Bendahan, Gautreau and Medina Escobar15 This expansion of per capita testing in a geographically distinct region enabled us to investigate how increased testing rates might impact test performance and influence CJD surveillance.
Of particular interest is the emergence of EP (and RT) – QuIC as a diagnostic tool offering enhanced reliability for the detection of sCJD, even in clinical scenarios characterized by a low pre-test probability. Typical sCJD (MM1 and MV1 subtypes) are the most commonly encountered forms in the clinic, accounting for 55% of cases detected in Canada following post-mortem analysis and are more sensitive to 14-3-3γ and total tau biomarker tests compared to other atypical CJD subtypesReference Slota, Myskiw and Peterson10,Reference Gmitterová, Heinemann and Bodemer16–Reference Bentivenga, Gonzalez-Ortiz and Baiardi19 which may also mimic other neurological conditions and complicate diagnosis. As EP-QuIC exhibits high sensitivity across sCJD subtypes, we reasoned that a focused region with increased EP-QuIC testing density may capture a greater number of atypical sCJD cases compared to regions with more typical testing patterns. Of vital importance is to determine whether increased testing of individuals with a low suspicion of CJD would result in increased false-positive test results. Therefore, here we examined whether the unusually high levels of EP-QuIC testing in NB were associated with altered patient demographics, 14-3-3γ and total tau biomarker levels, or sCJD molecular subtypes.
Results
Patient demographics are unique in NB compared to the rest of Canada
Between January 2016 and December 2024, 309 CSF samples originating from NB were submitted to the NML for analysis, a submission rate of 36.9 per million annually (age-adjusted; 95% CI: 32.8 to 41.4), significantly higher than any other province. The national submission rate (including NB) over the same period was 7.23 per million annually (age-adjusted; 95% CI: 6.96 to 7.51). Quarterly CSF sample submission rates for each province across the study period are illustrated in Figure 1. We previously noted that the increased sample submission rate in NB was associated with a higher but not statistically significant increase in CJD incidence compared to rest of Canada (2.51 and 1.51 annual positive tests per million, respectively).Reference Slota, Myskiw and Peterson10 Nationally, test results from 2743 CSF samples from unique individuals were reported during the study period resulting in 2741 EP-QuIC tests, 2676 14-3-3γ and 2671 total tau tests. Test results from CSF samples disaggregated by province are shown in Table 1. In some instances, CSF samples were associated with genetic testing of the PRNP open reading frame. 38 samples with silent or non-causative mutations were included in the analysis (A117A [23], 1OPRD [6], P68P [2], G124G [2], P39Q [1], G54S [1], E219K [1] and 1OPRD/I251T [1]), and 25 with causative mutations were excluded (E200K [13], V210I [2], P105S [2], D178N [2], 5OPRI [2], T183S [1], P102L [1], E196K [1] and 4OPRD [1]).
Quarterly CSF submissions by province between January 1st 2016 and December 31st 2025. Bars represent age-adjusted submissions per province, broken down quarterly. CSF = cerebrospinal fluid.
Figure 1. Long description
A stacked bar graph comparing the number of cerebrospinal fluid (CSF) submissions per million by province in Canada from 2016 to 2025. The horizontal axis represents the quarters from 2016 to 2025, and the vertical axis represents the number of CSF submissions per million. Each bar is divided into segments representing different provinces, with each segment color-coded according to the legend on the right. The provinces include Alberta (AB), British Columbia (BC), Manitoba (MB), New Brunswick (NB), Newfoundland and Labrador (NL), Nova Scotia (NS), Ontario (ON), Prince Edward Island (PE), Quebec (QC), and Saskatchewan (SK). The graph shows variations in the number of submissions over time, with notable peaks and troughs. The highest peak occurs in the first quarter of 2022, with significant contributions from several provinces. The data includes both actual and forecasted submissions, with the forecasted data extending to 2025.
Summary of cerebrospinal fluid test results by province
Table 1. Long description
The table presents data on cerebrospinal fluid test results across various provinces. It includes columns for province, population, and test results for EP-QuIC, 14-3-3, and total tau. The EP-QuIC column is further divided into Positive, Negative, and Indeterminate results. The 14-3-3 and total tau columns are divided into Positive and Negative results. The table has 11 rows, each representing a different province, and 10 columns. Row 1: AB, 4.45, EP-QuIC: Positive 49, Negative 150, Indeterminate 0, 14-3-3: Positive 56, Negative 137, Total tau: Positive 65, Negative 128. Row 2: BC, 5.27, EP-QuIC: Positive 80, Negative 311, Indeterminate 5, 14-3-3: Positive 88, Negative 295, Total tau: Positive 114, Negative 268. Row 3: MB, 1.40, EP-QuIC: Positive 23, Negative 63, Indeterminate 0, 14-3-3: Positive 22, Negative 62, Total tau: Positive 30, Negative 54. Row 4: NB, 0.80, EP-QuIC: Positive 21, Negative 287, Indeterminate 1, 14-3-3: Positive 31, Negative 273, Total tau: Positive 41, Negative 263. Row 5: NL, 0.53, EP-QuIC: Positive 9, Negative 16, Indeterminate 0, 14-3-3: Positive 9, Negative 16, Total tau: Positive 11, Negative 14. Row 6: NS, 1.01, EP-QuIC: Positive 11, Negative 35, Indeterminate 1, 14-3-3: Positive 13, Negative 33, Total tau: Positive 15, Negative 31. Row 7: ON, 14.94, EP-QuIC: Positive 186, Negative 650, Indeterminate 19, 14-3-3: Positive 218, Negative 611, Total tau: Positive 242, Negative 584. Row 8: PE, 0.16, EP-QuIC: Positive 1, Negative 2, Indeterminate 0, 14-3-3: Positive 2, Negative 1, Total tau: Positive 3, Negative 0. Row 9: QC, 8.60, EP-QuIC: Positive 171, Negative 566, Indeterminate 11, 14-3-3: Positive 185, Negative 553, Total tau: Positive 216, Negative 521. Row 10: SK, 1.17, EP-QuIC: Positive 12, Negative 61, Indeterminate 0, 14-3-3: Positive 19, Negative 52, Total tau: Positive 20, Negative 51. Row 11: Total, 38.32, EP-QuIC: Positive 563, Negative 2141, Indeterminate 37, 14-3-3: Positive 643, Negative 2033, Total tau: Positive 757, Negative 1914.
EP-QuIC = end-point quaking-induced conversion; AB = Alberta; BC = British Columbia; MB = Manitoba; NB = New Brunswick; NL = Newfoundland; NS = Nova Scotia; ON = Ontario; PE = Prince Edward Island; QC = Quebec; SK = Saskatchewan. Population values are expressed in millions. Data reflect 2021 estimates.
We assumed that given the high rate of samples submitted by NB, and the comparable CJD incidence to other provinces, it was likely that the submission criteria for CSF testing would differ between NB and other centers, which would translate into differences in sample demographics. Since CJD appeared to be high on the differential diagnosis during investigation of case evaluations of the NSUC, we reasoned that increased submissions likely include individuals from this cohort. A comparison of patient demographics for the CSF samples submitted from NB between 2016 and 2024 showed distinct differences from that of the rest of Canada. In particular, we observed that among QuIC-negative CSF samples, samples submitted from NB were clearly associated with lower patient age than most other Canadian provinces (Figure 2A). We also did not observe any significant differences in age among QuIC-positive CSF submissions, suggesting that EP-QuIC-positive cases in NB still occurred in individuals of a typical demographic range (Figure 2B). No significant differences were observed in the ratio of male to female samples submitted (Figure 2C).
Age distribution samples submitted for CSF analysis disaggregated by provinces. Violin plots show age distribution associated with CSF samples submitted among EP-QuIC-negative (panel A) and EP-QuIC-positive (panel B) samples. Overlaid box plots indicate the median age (horizontal line) and interquartile range (IQR; box boundaries), and vertical lines extend to 1.5 × IQ. Lower panels show mean ages per province, with error bars showing Wilson confidence intervals. Panel C shows province-level estimates of the female test proportion for all CSF samples, derived from a generalized linear model. Error bars represent 95% confidence intervals obtained by estimated marginal means. CSF = cerebrospinal fluid; EP-QuIC = end-point quaking-induced conversion.
Surrogate biomarker levels differ significantly across provinces within EP-QuIC-negative samples
We next compared the levels of the biomarkers 14-3-3γ and total tau across provinces. These markers represent proteins that are released from degenerating neurons and are therefore highly indicative of the rapid neurodegeneration seen in classic sCJD,Reference Zerr, Bodemer and Otto20–Reference Kapaki, Kilidireas, Paraskevas, Michalopoulou and Patsouris22 but are by no means specific for CJD and false negatives can occur. 14-3-3γ can also be detected in other causes of rapid dementia, including autoimmune encephalitis and neurodegenerative diseasesReference Kong, Chen, Shi, Zuo and Zhang23 and elevations in total tau are seen in conditions of synaptic degeneration, such as Alzheimer disease.Reference Soares, Bellaver and Ferreira24 Until recently (2025), all CSF samples submitted for EP-QuIC analysis were measured in parallel for 14-3-3γ and total tau levels.
Among QuIC-negative CSF samples, those submitted by NB were associated with the lowest 14-3-3γ and total tau levels of any province (Figure 3A, B). Contrarily, among QuIC-positive CSF samples submitted, no significant differences in the levels of either biomarker were found across provinces (Figure 3C, D). These findings support the notion that underlying demographic differences may exist across EP-QuIC-negative samples submitted from NB and could contribute to the observed variation in biomarker levels.
Biomarker abundance in CSF samples across provinces stratified by EP-QuIC status. Violin plots show the distribution of Log1p-transformed 14-3-3γ (panel A) and total tau (panel B) abundances within EP-QuIC-negative CSF samples. Overlaid box plots indicate the median biomarker levels (horizontal line) and interquartile range (IQR; box boundaries), and vertical lines extend to 1.5 × IQ. Lower panels show mean abundance per province, with error bars showing Wilson confidence intervals derived by linear regression analysis. Panels C and D show the same analysis performed within EP-QuIC positive CSF samples. CSF = cerebrospinal fluid; EP-QuIC = end-point quaking-induced conversion.
Test performance metrics vary among samples submitted from different provinces
To determine if the posited demographic differences introduced through increased test rates affects the predictive value of the biomarkers 14-3-3γ and total tau, we calculated biomarker test performance across provinces using our standard laboratory thresholds. We used EP-QuIC results as a surrogate for true prion disease status, given the specificity of the test is close to 100%.Reference Slota, Myskiw and Peterson10 Sensitivity, specificity, and positive/negative predictive values (PPV;NPV) for both biomarker tests were calculated using EP-QuIC status as a reference. Test metrics across provinces are shown in Table 2 and Table 3. Excluding PE, the PPV of both 14-3-3γ and tau tests were lowest in NB (0.50, 95% CI: 0.33 to 0.67 and 0.48, 95% CI: 0.33 to 0.63, respectively), which also had the second lowest 14-3-3γ test sensitivity behind MB (0.71; 95% CI: 0.50 to 0.86) and highest 14-3-3γ and total tau test specificity of all provinces (0.95, 95% CI: 0.91 to 0.97). Similar trends were observed with total tau levels, where NB had the lowest PPV (0.48, 95% CI: 0.33 to 0.63), and highest test specificity (0.93, 95% CI: 0.89 to 0.95) in the country.
14-3-3γ test performance by province
Table 2. Long description
A table comparing test performance metrics for 14-3-3 by province. The table has 10 rows and 8 columns. Columns are labeled TN, TP, FN, FP, Sensitivity, Specificity, PPV, and NPV. Rows are labeled with province abbreviations: AB, BC, MB, NB, NL, NS, ON, PE, QC, and SK. Row 1: AB, TN 127, TP 37, FN 10, FP 19, Sensitivity 0.79 [0.65-0.88], Specificity 0.87 [0.81-0.92], PPV 0.66 [0.53-0.77], NPV 0.93 [0.87-0.96]. Row 2: BC, TN 272, TP 59, FN 19, FP 28, Sensitivity 0.76 [0.65-0.84], Specificity 0.91 [0.87-0.93], PPV 0.68 [0.57-0.77], NPV 0.93 [0.90-0.96]. Row 3: MB, TN 54, TP 15, FN 8, FP 7, Sensitivity 0.65 [0.45-0.81], Specificity 0.89 [0.78-0.94], PPV 0.68 [0.47-0.84], NPV 0.87 [0.77-0.93]. Row 4: NB, TN 267, TP 15, FN 6, FP 15, Sensitivity 0.71 [0.50-0.86], Specificity 0.95 [0.91-0.97], PPV 0.50 [0.33-0.67], NPV 0.98 [0.95-0.99]. Row 5: NL, TN 14, TP 7, FN 2, FP 2, Sensitivity 0.78 [0.45-0.94], Specificity 0.88 [0.64-0.97], PPV 0.78 [0.45-0.94], NPV 0.88 [0.64-0.97]. Row 6: NS, TN 30, TP 9, FN 2, FP 4, Sensitivity 0.82 [0.52-0.95], Specificity 0.88 [0.73-0.95], PPV 0.69 [0.42-0.87], NPV 0.94 [0.80-0.98]. Row 7: ON, TN 553, TP 140, FN 43, FP 74, Sensitivity 0.77 [0.70-0.82], Specificity 0.88 [0.85-0.90], PPV 0.65 [0.59-0.71], NPV 0.93 [0.90-0.95]. Row 8: PE, TN 1, TP 1, FN 0, FP 1, Sensitivity 1.00 [0.21-1.00], Specificity 0.50 [0.09-0.91], PPV 0.50 [0.09-0.91], NPV 1.00 [0.21-1.00]. Row 9: QC, TN 509, TP 133, FN 36, FP 50, Sensitivity 0.79 [0.72-0.84], Specificity 0.91 [0.88-0.93], PPV 0.73 [0.66-0.79], NPV 0.93 [0.91-0.95]. Row 10: SK, TN 51, TP 11, FN 1, FP 8, Sensitivity 0.92 [0.65-0.99], Specificity 0.86 [0.75-0.93], PPV 0.58 [0.36-0.77], NPV 0.98 [0.90-1.00].
TN = true negative; TP = true positive; FN = false negative; FP = false positive; PPV = positive predictive value; NPV = negative predictive value; AB = Alberta; BC = British Columbia; MB = Manitoba; NB = New Brunswick; NL = Newfoundland; NS = Nova Scotia; ON = Ontario; PE = Prince Edward Island; QC = Quebec; SK = Saskatchewan. Sensitivity = TP/(TP + FN); Specificity = TN/(TN + FP); PPV = TP/ TP + FP); NPV = TN/(TN + FN). Numbers in brackets are confidence intervals calculated using the Wilson score method for binomial proportions.
Total tau test performance by province
Table 3. Long description
The table presents biomarker test performance metrics across various provinces. It includes columns for true Negatives (TN), true Positives (TP), false Negatives (FN), false Positives (FP), Sensitivity, Specificity, Positive Predictive Value (PPV), and Negative Predictive Value (NPV). The provinces listed are AB, BC, MB, NB, NL, NS, ON, PE, QC, and SK. Each row provides specific values for these metrics in each province. For example, in AB, TN is 124, TP is 43, FN is 4, FP is 22, Sensitivity is 0.91 [0.80-0.97], Specificity is 0.85 [0.78-0.90], PPV is 0.66 [0.54-0.76], and NPV is 0.97 [0.92-0.99]. Similar detailed values are provided for each province, showing variations in test performance across different regions.
TN = true negative; TP = true positive; FN = false negative; FP = false positive; PPV = positive predictive value; NPV = negative predictive value; AB = Alberta; BC = British Columbia; MB = Manitoba; NB = New Brunswick; NL = Newfoundland; NS = Nova Scotia; ON = Ontario; PE = Prince Edward Island; QC = Quebec; SK = Saskatchewan. Sensitivity = TP / (TP + FN); Specificity = TN / (TN + FP); PPV = TP / (TP + FP); NPV = TN / (TN + FN). Numbers in brackets are confidence intervals calculated using the Wilson score method for binomial proportions.
Post-mortem data associated with NB cases
Of the 21 EP-QuIC positive cases from NB, 19 underwent neuropathological examination and/or Western blot analysis. 16/18 were positive by Western blot upon examination of frozen sections taken from the frontal cortex, striatum, and cerebellum. 19/19 were positive by neuropathological examination indicating 100% specificity of EP-QuIC in this cohort. 3 EP-QuIC cases were not evaluated post-mortem. The single indeterminate EP-QuIC case was also confirmed as CJD by both Western blot and neuropathological assessment. Since the age distribution and biomarker test results suggested an altered patient demographic in the NB cohort relative to the remainder of Canada, we reasoned that this might contribute to different glycoform profiles in CJD cases from the province. We evaluated the protease-resistant PrP glycoforms across all provinces by Western blot. The proportion of sporadic CJD molecular subtypes detected across Canadian provinces is reported in Figure 4. Within the 16 EP-QuIC-positive or indeterminate cases from NB that tested positive by Western blot, we detected 4 MM1 (25%), 4 MV1 (25%), 6 VV2 (37%), 1 MM2 (6%) and 1 mixed MV case (6%; Figure 1). 1 of the 6 VV2 samples was associated with an indeterminate CSF EP-QuIC test. Although the total number of autopsy-confirmed cases are low in NB, it is of note that a lower proportion of VV2 cases were detected in AB, BC, MB, ON and QC (17%, 14%, 6%, 17% and 21%, respectively). The relative abundance VV2 subtypes observed in these provinces are more in line with what has been reported elsewhere.Reference Parchi, Giese and Capellari25 Although case numbers are too low to allow any statistical significance to be attached to this observation, the trend could be indicative of an increased capacity to identify and diagnose type 2 CJD cases in NB.
Prion disease subtype distribution across provinces throughout the study period. Relative abundance of indicated prion subtypes is shown per province for all cases sent to autopsy throughout the study period. Numbers above bars indicated sample numbers (n) per province.
Discussion
Specific diagnosis of CJD using testing based on QuIC technology has revolutionized the ability to accurately identify probable cases prior to death and neuropathological analysis of brain tissue. These tests have been validated in several countries over the last 10 years, with the specificity determined to be close to 100% for the EP-QuIC test used in Canada.Reference Slota, Myskiw and Peterson10 Despite this impressive specificity, we are aware that CSF samples are usually submitted for testing in those cases that are reasonably suspected of being CJD, based on the clinical case definition for classic CJD. Given that 15–20% of CJD cases have atypical presentations, many of which closely mimic other conditions, the case can be made for more widespread use of CJD diagnostic testing as part of CSF panel testing for various conditions in cases with no clear diagnosis. However, this strategy has some inherent risk; a false positive test result indicating a fatal, untreatable condition such as CJD would impact patient care and well-being. No studies exploring the use of more widespread testing, or screening, of patients with undiagnosed neurological syndromes have been performed to determine the utility of QuIC in such a cohort. In this study, we compared CSF samples across Canadian provinces that were submitted to our laboratory for QuIC testing. NB had over 5 times the per capita submissions compared to the rest of Canada over the study period, 2016–2024. Many of these samples likely originated from patients with neurological symptoms, some of which may mimic CJD or other conditions, based on the case definitions published in relation to the NSUCReference Bendahan, Gautreau and Medina Escobar15 which includes a broader range of neurological symptoms than those used to identify suspected cases of classic CJD. The case definition used to assign confirmed and suspect cases of NSUC can be found as Supplementary data in the manuscript of Bendahan et al., 2025.Reference Bendahan, Gautreau and Medina Escobar15
Our findings indicate that increased CSF testing rates in NB expanded the patient cohort to include those outside of the typical demographic that are tested for CJD. Although the diagnostic lab does not have access to the clinical findings that prompted a CSF test requisition, this is supported by observations of lower surrogate biomarker levels and younger patient demographics concurrent with our previous observation of a slight but insignificant increase in CJD incidence in the province estimated using EP-QuIC positive tests.Reference Slota, Myskiw and Peterson10 The fact that no false positive tests were detected in this cohort suggests that EP-QuIC can be broadly and reliably applied in the clinic even in cases where clinical profiles may fall outside of the typical MM(V)1 spectrum. This may ultimately serve to improve detection of atypical CJD cases, including slow-progressing MM/MV2C subtypes that pose a challenge to identify based on clinical findings alone.Reference Hamaguchi, Sanjo and Ae26,Reference Baiardi, Vargiu and Appleby27 The most common prion disease subtypes MM1 and VV2 have relative prevalences of ∼40% and 20%, respectively, as reported by Canada.Reference Slota, Myskiw and Peterson10 and elsewhere.Reference Parchi, Strammiello and Notari28,Reference Rossi, Baiardi and Parchi29 Although the total number of Western blot positive cases in NB analyzed here were limited to 16,, we noticed a slight increase in VV2 cases from NB. VV2 CJD cases can present with clinical features that deviate from the established diagnostic criteria for classic CJD, complicating their early diagnosis.Reference Baiardi, Magherini and Capellari30 Because EP-QuIC is reliable across prion disease subtypes,Reference Slota, Myskiw and Peterson10 intensive regional CSF testing may have facilitated the identification of atypical CJD cases that might otherwise have gone undetected in NB.
We also assessed the performance of the total tau and 14-3-3γ tests using EP-QuIC test status as a gold standard. Until 2025, CSF testing for prion diseases in Canada consisted of administering EP-QuIC, 14-3-3γ and tau tests for all samples, and CSF samples are accepted agnostic to any tests. The PPV provides the likelihood that a positive test corresponds to positive disease status in the tested population and can be confounded in instances where disease prevalence is low (the case for exceedingly rare diseases such as CJD), and false positive test results can approach the number of true positive cases. Again, we attribute this to an expanded umbrella of clinical conditions under which CSF samples were submitted for CJD testing, effectively reducing the disease prevalence in the tested population. Interestingly, the low levels of total tau and 14-3-3γ in the NB cohort suggest that the referred cases have a low level of demonstrable neuronal degeneration.
Lastly, these findings may help to inform surveillance efforts. As programs shift from focused efforts aimed at detecting variant CJD, to detecting emerging prion diseases such as chronic wasting disease (CWD),Reference Watson, Brandel and Green31 strategies to detect novel prion diseases with atypical clinical presentations will be important. Focused testing in CWD-endemic regions should rely on prion-specific tests such as EP-QuIC, as well as improved protocols that increase detection sensitivity while preserving high specificity.Reference Myskiw, Fox and Kielich8 These tests report directly and reliably on the presence of the disease-causing agent and will be essential tools to support future surveillance efforts aimed at intensive testing of focused populations or the introduction of more expansive prion testing strategies.
Methods
Cohort description
Patient-level results for CSF biomarker and EP-QuIC tests, genetic testing and biochemical subtyping were extracted from the NML Laboratory Information System (LIMS) and filtered to include those with CSF sample collection dates between January 1st, 2016 and December 31, 2024. Samples where CSF was bloody or xanthochromic were excluded from the cohort, as were the genetic cases as indicated in the main text. Patient age was calculated using the date of CSF collection and patient date of birth recorded at the time of sample submission. Patient sex was retrieved from the test requisition form submitted with test samples.
Laboratory testing
Test results for each CSF sample were extracted from LIMS using standard laboratory test thresholds for the reported biomarker levels. 14-3-3γ levels were measured using the Circulex 14-3-3γ ELISA kit, and total tau levels were measured using the Innotest hTAU ELISA kit according to the manufacturer’s instructions. Test thresholds of 20,000 absorbance units, or 976 pg/mL, were used for 14-3-3γ and total tau tests, respectively, to determine positivity. EP-QuIC was performed as described previously using in-house purified full-length Syrian golden hamster PrP substrate (residues 23-231).Reference Cheng, Vendramelli and Sloan7
Data analysis
To calculate age-adjusted CSF test incidences, CSF samples were separated into the following age strata according to patient age at the time of sample collection: 0–17 years, 18–24 years, 25–44 years, 45–64 years and >64 years. Age-stratified CSF samples were associated with annual province-level population data retrieved from Statistics Canada to calculate per-province age-specific incidences. Age-adjusted incidences and exact confidence intervals were calculated by direct method with the epitools R package,Reference Fay and Feuer32 using reference weights derived from the 2011 Canadian population. Differences in biomarker levels across provinces were assessed using a generalized linear model (glm) in R, with biomarker level as the response variable and province as the predictor. Model errors were estimated assuming a gamma distribution with a log-link function. Pairwise comparisons between provinces were conducted using estimated marginal means calculated with the emmeans package in R. Post hoc significance testing was performed using Tukey’s method. To analyze patient age across provinces, a linear model was used. Post hoc pairwise comparisons were performed using the same approach used for biomarker analyses. Disease incidence and testing rates were analyzed using a glm in R, with mean estimates calculated using a log-linked Poisson distribution. Calculation of estimated marginal means and post hoc significance testing were performed as described above. Statistical outputs are provided in a Supplementary File 1.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/cjn.2026.10674.
Acknowledgements
We would like to sincerely thank the patients from whom diagnostic specimens were collected, as well as their families. We also acknowledge the physicians and clinical staff involved in patient management and sample collection. In addition, we recognize the contributions of hospital laboratories and provincial public health laboratories across Canada for providing clinical specimens and associated data used in this study. We also acknowledge colleagues at the Centre for Food-borne, Environmental and Zoonotic Infectious Diseases for their role in coordinating the autopsy program.
Author contributions
B.A.B.E and S.A.B. designed the study. J.M., K.A., C.P., D.M.S.K., L.L., K.F. and S.L.R.S. performed diagnostic testing and collated data. G.H.J. performed pathological analyses and case review. B.A.B.E. analyzed data and prepared figures. B.A.B.E. drafted the manuscript. B.A.B.E., J.A.S., V.L.S. and S.A.B. contributed to writing, review and editing.
Funding statement
This work was funded by the Public Health Agency of Canada.
Competing interests
Dr Valerie Sim is contracted by PHAC for clinical neurology services. The authors declare no other competing interests.
Open access
