Premium
Prediction of amyloid PET status using the LUMIPULSE G β‐amyloid ratio (1‐42/1‐40)
Author(s) -
Willemse Eline A.J.,
Tijms Betty M.,
Van Berckel Bart N.M.,
van Der Flier Wiesje,
Gan Sara,
Bastard Nathalie Le,
Radwan Rachel R.,
Esquivel Rianne N.,
Latham Jessica,
Benitalya,
Dickson Diana,
Scheltens Philip,
Teunissen Charlotte E.
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.046006
Subject(s) - medicine , concordance , dementia , biomarker , cohort , cognitive impairment , youden's j statistic , cutoff , cerebrospinal fluid , amyloid (mycology) , gastroenterology , nuclear medicine , pathology , receiver operating characteristic , disease , biochemistry , chemistry , physics , quantum mechanics
Background Cerebrospinal fluid (CSF) amyloid biomarker concentrations are valuable in the assessment of patients for Alzheimer’s dementia (AD). Automated platforms and the additional measurement of Aβ1‐40(Aβ40) with Aβ1‐42(Aβ42) has decreased preanalytical variability and increased diagnostic accuracy (Delaby, 2019; Hansson, 2019). Notably, the ratio Aβ42/Aβ40 significantly improves concordance with PET when compared to Aβ42 alone (Hansson, 2019). In this study Aβ42/Aβ40 concordance with PET was evaluated on the LUMIPULSE G1200 platform. Methods A subset of biobanked individuals from the Amsterdam Dementia Cohort was included based on availability of visual amyloid‐PET reads using an FDA approved tracer (Florbetaben n=156; Flutemetamol n=71; Florbetapir n= 8) and CSF, and an age requirement of ≥50 years old (n=235). Demographics included 58% amyloid‐PET positive; mean age 63 years; 43% female; MMSE range 12‐30; and diagnoses: subjective cognitive decline (n=53), mild cognitive impairment (n=25), AD (n=105), or other dementias (n=52). CSF Aβ42 and Aβ40 were measured using Lumipulse G1200 (Fujirebio Diagnostics Inc., Malvern, PA). A cutoff for Aβ42/Aβ40 was established by maximizing the Youden index subject with constraint for sensitivity and specificity to differ by no more than 10 percentage points. Results At a set cutoff of 646 pg/mL CSF Aβ42 sensitivity and specificity was 78.8% (CI 71.3% – 84.8%) and 80.6% (CI 71.7% – 87.2%) with an AUC of 0.854. CSF Aβ40 sensitivity and specificity was 51.8% (CI 43.5% – 60.0%) and 45.9% (CI 36.4% ‐ 55.8%) with an AUC of 0.401. The ratio Aβ42/Aβ40 at a set cutoff of 0.058 had a sensitivity and specificity of 94.2% (CI 88.9% ‐ 97.0%) and 90.8% (CI 83.5% and 95.1%) with an AUC of 0.952. Frequencies of patients within the AT(N) research framework were evaluated. With Aβ42 alone 127 patients were considered amyloid positive while use of the Aβ42/Aβ40 ratio included 138 amyloid positive patients. Conclusion The Aβ42/Aβ40 ratio strongly correlates with PET status when measured on the LUMIPULSE G platform. In conjunction with clinical assessment the Aβ42/Aβ40 ratio can serve as a highly sensitive and specific biomarker to assess amyloid status.