Blood‐brain barrier integrity impacts the use of plasma amyloid‐β as a proxy of brain amyloid‐β pathology

Abstract

INTRODUCTIONAmyloid‐β (Aβ) and tau can be quantified in blood. However, biological factors can influence the levels of brain‐derived proteins in the blood. The blood‐brain barrier (BBB) regulates protein transport between cerebrospinal fluid (CSF) and blood. BBB altered permeability might affect the relationship between brain and blood biomarkers.METHODSWe assessed 224 participants in research (TRIAD, n = 96) and clinical (BIODEGMAR, n = 128) cohorts with plasma and CSF/positron emission tomography Aβ, p‐tau, and albumin measures.RESULTSPlasma Aβ42/40 better identified CSF Aβ42/40 and Aβ‐PET positivity in individuals with high BBB permeability. An interaction between plasma Aβ42/40 and BBB permeability on CSF Aβ42/40 was observed. Voxel‐wise models estimated that the association of positron emission tomography (PET), with plasma Aβ was most affected by BBB permeability in AD‐related brain regions. BBB permeability did not significantly impact the relationship between brain and plasma p‐tau levels.DISCUSSIONThese findings suggest that BBB integrity may influence the performance of plasma Aβ, but not p‐tau, biomarkers in research and clinical settings.Highlights BBB permeability affects the association between brain and plasma Aβ levels. BBB integrity does not affect the association between brain and plasma p‐tau levels. Plasma Aβ was most affected by BBB permeability in AD‐related brain regions. BBB permeability increases with age but not according to cognitive status.

Fully automated measurement of plasma Aβ42/40 and p‐tau181: Analytical robustness and concordance with cerebrospinal fluid profile along the Alzheimer's disease continuum in two independent cohorts

“…The highest agreement between CSF and plasma biomarkers – both Aβ42/40 and p‐tau181 – was observed in patients with the highest CSF/serum albumin ratio. The difference between the highest CSF/serum albumin ratio tertile and the lower ones was statistically significant for Aβ42/40, which supports the influence of BBB permeability on the concordance between specific CSF and plasma biomarkers 12 …”

“…On the contrary, impaired renal function will likely generate false positive results on plasma p-tau181. 11 Following the results obtained by Bellaver et al, 12 we considered BBB permeability (reflected by CSF/serum albumin ratio levels) as another potential confounder. The highest agreement between CSF and plasma biomarkers -both Aβ42/40 and p-tau181 -was observed in patients with the highest CSF/serum albumin ratio.…”

“…The difference between the highest CSF/serum albumin ratio tertile and the lower ones was statistically significant for Aβ42/40, which supports the influence of BBB permeability on the concordance between specific CSF and plasma biomarkers. 12 Finally, we calculated probability profiles where specific values of plasma markers corresponded to predefined A/T categories in CSF.…”

“…The presence of KD was defined if the glomerular filtration rate was < 60 mL/min/1.73 m 2 11 . With respect to CSF/serum albumin ratio, patients were categorized into those with low, medium, and high ratio based on the tertile partition of the cohort as previously described by Bellaver et al 12 . Detailed information about how patients with AD, PD, PD with dementia (PDD), DLB, and FTD were diagnosed is present in the Supplementary Material.…”

“…The presence of KD was defined if the glomerular filtration rate was < 60 mL/min/1.73 m 2 . 11 With respect to CSF/serum albumin ratio, patients were categorized into those with low, medium, and high ratio based on the tertile partition of the cohort as previously described by Bellaver et al 12 Detailed information about how patients with AD, PD, PD with dementia (PDD), DLB, and FTD were diagnosed is present in the Supplementary Material. The UNIPG cohort was used to determine the distribution of plasma Aβ42/40 and p-tau181 in the different A/T categories and clinical groups, to evaluate the impact of KD and BBB permeability on the concordance between CSF and plasma biomarkers, and for model training.…”