Differential sensitivity of CSF markers of p-tau and t-tau to in vivo tau deposition assessed with [18F]-AV-1451 in typical and atypical Alzheimer’s disease

Objectives: To gain a better understanding of the relationship between CSF markers and PET imaging of tau pathology, we examined the association of tau deposition to phosphorylated-tau (p-tau) and total-tau (t-tau) in typical and atypical cases of Alzheimer’s disease (AD).

Methods: Fifteen typical and ten atypical AD patients (six patients with logopenic variant of primary progressive aphasia, two posterior cortical atrophy patients, one with behavioral dysexecutive variant, and one atypical patient with multiple abnormalities in clinical appearance) underwent a routine lumbar puncture where 20-30 ml cerebrospinal fluid was collected. Diagnosis of dementia of the Alzheimer’s type was based on CSF biomarker profile, MRI structural evaluation and clinical presentation of the patients. CSF samples were analyzed using an immunosorbent essay (INNOTEST ELISA; Fujirebio Europe). Additionally, all patients underwent a PET scan with [18F]-AV-1451, which was normalized to the non-specific tracer uptake in the cerebellum to obtain standard uptake value ratios (SUVR) images. First, we examined if CSF levels of p-tau and t-tau differed between typical and atypical AD and if differential relationships to in vivo global cortical tau burden are present. Second, we extracted SUVRs from cortical and subcortical regions, and estimated for each disease category separately, which regions best predicted CSF levels of p-tau and t-tau. These analyses were conducted within the framework of forward regression models controlling for the time lag between CSF assessment and PET scan as well as the age of the patient.

Results: CSF levels of p-tau and t-tau were not significantly different between typical and atypical AD patients (see Figure 1a). Interestingly, global cortical tau burden was positively associated with p-tau but not t-tau in typical AD, whereas in atypical AD no relationship of global tau burden and CSF levels were observed. The regional analysis revealed that in patients with typical AD, both CSF markers were associated with inferior parietal and temporal tau pathology, whereas p-tau was found to additionally relate to increased tau burden in posterior cingulum (see Figure 1b). T-tau levels in typical AD were associated with Heschl gyri and inferior temporal tau burden. In atypical AD, frontal regions were associated with both p-tau and t-tau, whereas differential sensitivity for p-tau was found with tau burden in the pallidum and for t-tau measures in the temporal pole.

Conclusion: Although both CSF-markers of p-tau and t-tau do not distinguish between AD types, global in vivo tau deposition and p-tau showed a selective association in typical but not atypical AD. Although preliminary, this result suggests that CSF tau marker in atypical AD do not reflect the extent of the pathological tau burden. We additionally identified a distinct spatial signature of each marker to in vivo regional tau burden in typical and atypical AD which may relate to the generally different distribution of tau deposits in different disease entities. We conclude that in vivo tau imaging can differentiate typical from atypical AD and allow better insights into the neurodegenerative spread of the disease, compared to CSF markers of tau. Research Support: NA.

• Download figure
• Open in new tab
• Download powerpoint