Skip to main content
. 2019 Mar 26;46(6):1276–1286. doi: 10.1007/s00259-019-04297-5

Table 2.

Agreement between CSF positivity and [18F]flutemetamol PET positivity using dichotomized measures based on the diagnoses made before [18F]flutemetamol PET

CSF [18F]Flutemetamol PETa CSF biomarkerb Number (%) of patients positive
MCI (n = 86c) AD (n = 38c) Non-AD disorder (n = 9c) Dementia NOS (n = 15c) SCD (n = 4c)
Positive Positive 1-42 25 (29) 19 (50) 2 (22) 2 (13)
p-tau 30 (35) 19 (50) 2 (22) 2 (13) 3 (75)
t-tau 29 (34) 20 (52) 3 (33) 2 (13) 3 (75)
Negative 1-42 11 (13) 3 (8) 3 (33) 2 (13)
p-tau 3 (3) 4 (10) 2 (22) 3 (20)
t-tau 3 (3) 3 (8) 1 (11) 3 (20)
Negative Positive 1-42 23 (27) 6 (16) 3 (75)
p-tau 18 (21) 6 (16)
t-tau 19 (22) 6 (16)
Negative 1-42 27 (31) 10 (26) 4 (45) 11 (74) 1 (25)
p-tau 35 (41) 9 (24) 5 (56) 10 (67) 1 (25)
t-tau 35 (41) 9 (24) 5 (56) 10 (67) 1 (25)

Concordance (both biomarkers positive or negative): CSF-positive/PET-positive and CSF-negative/PET-negative. Discordance (only one of two biomarkers positive): CSF-positive/PET-negative (isolated CSF positivity) and CSF-negative/PET-positive (isolated PET positivity)

aThe cut-off value used for [18F]flutemetamol SUVR was 0.60, in combination with visual assessment

bThe cut-off values used to binarize CSF biomarkers were <550 ng/L for Aβ1-42, >80 ng/L for p-tau and >400 ng/L for t-tau

cNumber of patients in whom CSF was sampled