HAI Book 2020 - Dec 18B

r’s disease, SV2A, PET
26

Min, Hoon-Ki

P151: Mental and physical activity du
target binding in extra- and within-br

Hoon-Ki (Paul) Min1, Christopher Apgar1, Nanc
Albertson1, Christopher Schwarz1, Hugo Botha1
Petersen1, Clifford Jack1, Val Lowe1

1Mayo Clinic, Rochester, MN, US
With tau PET imaging now available, there has been g
changes in Alzheimer’s Disease. Thus, identifying fac
critical. Off-target tau PET(18F-FTP) signals both wit
which might bleed into on-target tau regions. At the 2
that walking and sleeping activity during the uptake p
signal(n=136). The current study of n=330 subjects co
within-brain areas. In our standard tau PET procedure
during the uptake period. They may engage in various
questionnaire on uptake activities following their PET
brain areas were measured in SUV and SUVR units. L
relationship between the participant’s activities and lig
areas. Participants who reported ‘sleep’ during the upt
muscles and cortical bone/meninges, while participan
those areas (p<0.05). The effects of ‘sleep’ were relate
signal (p<0.05). For within-brain regions, only ‘walki
ganglia. Meta-ROI of on-target tau regions, WM refer
was not affected by uptake activities. We have observ
extra-brain off-target tau signals; (2) affect within-bra
brain on-target tau signals. Our results suggest implem
tau PET that would help reduce extra- and within-brai

42

uring the uptake period affects off-
rain Tau PET (18F-FTP)

cy Scott1, Haakon Hol1, Emily Lundt1, Sabrina
1, Prashanthi Vemuri1, Jeffrey Gunter1, Ronald

growing interest to accurately trace longitudinal tau
ctors that cause unwanted PET signal variability is
thin and external to the brain have been observed,
2019 HAI, we presented preliminary results showing
period affects the extra-brain off-target tau
onfirms our initial results and reviews effects in
e, following ligand infusion, we release the participant
s activities prior to scanning. We administered a
T scan. The tau values of off-target extra- and within-
Linear regression models were used to assess the
gand distribution in brain and non-brain
take period had higher SUV values in oculomotor
nts who reported ‘walking’ had a lower uptake value in
ed to time; the more sleep, the greater the off-target
ing’ was associated with decreased SUVR in the basal
rence regions and internal carotid artery (control area)
ved that certain activities during uptake: (1) affect
ain off-target tau signals; and (3) do not affect within-
menting an uptake activity restriction protocol during
in off-target signal variability.

27

42

28

Keywords: Tau, PET, Off-target, Sleep, Walk
42

29

Nakano, Yoshikazu

P152: In vivo uptake of 18F-PM-PBB3
corticobasal syndrome

Yoshikazu Nakano1, Hitoshi Shimada1, Kenji T
Keisuke Takahata1, Yuhei Takado1, Hitoshi Shi
Sano1, Chie Seki1, Shigeki Hirano1,3, Maiko On
Sahara1, Kazunori Kawamura1, Ming-Rong Zha

1National Institute of Radiological Sciences, National
Technology, Chiba, Japan
2Neurology Clinic Chiba, Chiba, Japan
3Department of Neurology, Chiba University Graduat
4APRINOIA Therapeutics Inc, Taipei, Taiwan

Objectives: Corticobasal syndrome (CBS) is a clinica
backgrounds chiefly consisting of Alzheimer’s disease
degeneration and progressive supranuclear palsy. Aim
background in CBS by PET scan with 18F-PM-PBB3
tau aggregates, and 3) to assess its correlation with cli
suspected non-AD tauopathy.

Methods: Seven patients diagnosed as CBS by Camb
(HCs) were recruited. The age and years of education
underwent PET scans with 18F-PM-PBB3 and 11C-Pi
Parametric PET images were generated by voxel-base
using the cerebellar cortex as a reference region. 18F-P
negative with visual assessment. Participants were cla
A(+)) by 11C-PiB PET and tau positive or negative((T
CBS cases, 18F-PM-PBB3 SUVR were compared with

Results: One of the seven CBS cases was amyloid po
PBB3 accumulation in A(+) case was obviously highe
CBS cases had predominant symptoms in their right s
accumulation in left globus pallidum and precentral gy
predilection for tau pathologies in CBD.

Conclusions: 18F-PM-PBB3 was able to track tau lesi
ligand may be applicable to diagnose tauopathy one b

Keywords: CBS, CBD, 18F-PM-PBB3 (18F-APN-167

43

3 (18F-APN-1607) in patients with

Tagai1, Kiwamu Matsuoka1, Manabu Kubota1,
inotoh1,2, Yasuharu Yamamoto1, Yasunori
no1, Paul Tempest4, Ming-Kuei Jang4, Naruhiko
ang1, Satoshi Kuwabara3, Makoto Higuchi1

l Institutes for Quantum and Radiological Science and

te School of Medicine, Chiba, Japan

al syndrome based on diverse pathological
e (AD) and 4-repeat tauopathies such as corticobasal
ms of the present study are 1) to estimate pathological
(18F-APN-1607), 2) to characterize the distribution of
inical features in patients with CBS based on

bridge’s criteria, and nine cognitively healthy controls
matched between two groups. All participants
iB for estimating regional tau and Aβ deposition.
ed calculation of standard uptake value ratio (SUVR)
PM-PBB3 and 11C-PiB uptake were judged positive or

assified as amyloid positive or negative ((A(-) or
T(+) or T(-)) by 18F-PM-PBB3 PET. As for A(-)T(+)
h HCs by voxel-base analysis.
ositive, and another case was tau negative. 18F-PM-
er than the other A(-)T(+) CBS cases. Five A(-)T(+)
side limbs, and the voxel-base analysis showed higher
yrus, which are pathologically reported as areas of

ions in patients with CBS. Visual assessment of this
by one.
70)

30

Ossenkoppele, Rik

P153: Factors predicting tau PET stat
impaired individuals

Rik Ossenkoppele1, Antoine Leuzy1, Hannah Ch
Matsson1, Sebastian Palmqvist1, Chul Lyoo3, Gi

1Lund University, Lund, Sweden
2UCSF, San Francisco, CA, US
3Gangam Severance Hospital, Seoul, Korea

Background: A substantial proportion of Aβ+ patient
(MCI) are tau PET-negative, while several individuals
neurodegenerative disorder are tau PET-positive. We
PET status in early to late Braak regions-of-interest (R

Methods: We included 1312 participants (457 cogniti
and 370 non-AD neurodegenerative disorder) who un
or [18F]RO948 (n=702) at research centers in Sweden,
positivity were previously determined in Braak I-II, B
mean+2*SD in older healthy controls for [18F]flortauc
for [18F]RO948. To identify factors predicting tau stat
models with tau-status in the Braak regions (+/-) as de
(APOE) ε4 status, Aβ-status (only in analyses includin
cortical thickness in an AD-signature region as predic
logistic regression models to account for all other pred

Results: Participant characteristics are presented in T

Table 1. Participant characteristics Aβ-

Aβ-positive AD
dementia

N 296

Age 70.8±8.9

Sex, %male 42.9

MMSE 20.2±5.0
APOE ε4+, % 57.8

Aβ+, % 100

[18F]FTP/[18F]RO948, n 179/117

Tau PET + Braak I/II, % 86.8

Tau PET + Braak I/IV, % 89.5

Tau PET + Braak V/VI, % 70.9

AD-signature thickness, mm3 2.35±0.17

In Aβ+ AD dementia, age (younger in Braak I-II and o
MMSE scores were associated with tau-negativity (Ta
higher cortical thickness were associated with tau-neg
individuals with normal cognition and non-AD neurod
lower MMSE scores and reduced cortical thickness w
(Table-3).

43

tus in cognitively unimpaired and

ho3, Renaud La Joie2, Olof Strandberg1, Niklas
il Rabinovici2, Ruben Smith1, Oskar Hansson1

ts with AD dementia and mild cognitive impairment
s with normal cognition or a non-AD
aimed to investigate which factors contribute to tau
ROIs).

ively normal, 189 Aβ+ MCI, 296 Aβ+ AD dementia
nderwent tau PET with either [18F]flortaucipir (n=676)
, South Korea and the USA. Cut-offs for tau-
Braak I-IV and Braak V-VI ROIs using the
cipir and the mean+2.5*SD in young healthy controls
tus, we performed bivariate binary logistic regression
ependent variable and age, sex, apolipoprotein
ng controls and non-AD participants), MMSE and
ctors. Additionally, we performed multivariable binary
dictors in the same model.

Table-1.

-positive MCI Non-AD disorders Cognitively
unimpaired
189 370
71.2±8.6 69.3±8.3 457
65.8±13.2
52.3 58.1
26.4±2.6 24.1±5.7 43.3
28.8±1.3
59.3 22.4
100 25.4 39.4
83/106 254/116 30.2
47.1 18.4 160/297
47.1 9.7 10.1
26.5 4.9 4.4
2.51±0.19 2.50±0.20 2.1
2.67±0.16

older in Braak I-IV and V-VI regions) and higher
able-2). In Aβ+ MCI, APOE ε4 non-carriership and

gativity in Braak I-IV and V-VI regions (Table-2). In
degenerative disorders, presence of Aβ pathology,

were the main predictors of tau-positivity across ROIs

31

Table 2. Factors contributing to T(au) status in Aβ+ A

Aβ+ AD de

Braak I-II

(T+ = 1, T- = 0)

OR (95% CI) P OR (95

A. UNIVARIATE MODEL

Age 1.04(1.01-1.07) <0.001 0.90(0.

Sex, % male 0.72(0.44-1.18) 0.188 0.90(0.
APOE ε4+, % 1.85(1.09-3.13) 0.023 2.57(1.

MMSE 0.96(0.91-1.01) 0.169 0.80(0.

Thickness 0.67(0.16-2.73) 0.577 0.67(0.

B. MULTIVARIABLE MODEL

Age 1.04(1.01-1.08) 0.006 0.88(0.

Sex, % male 0.95(0.54-1.66 0.845 1.32(0.
APOE ε4+, % 2.04(1.17-3.57) 0.012 2.44(1.

MMSE 0.96(0.90-1.02) 0.212 0.79(0.

Thickness 0.96(0.17-5.33) 0.964 0.18(0.

Aβ+ Mild cognitiv

OR (95% CI) P OR (95

A. UNIVARIATE MODEL

Age 1.04(1.01-1.08) 0.028 1.01(0.

Sex, % male 1.03(0.58-1.83) 0.911 0.67(0.
APOE ε4+, % 1.88(1.00-3.55) 0.050 1.86(0.

MMSE 0.96(0.86-1.08) 0.525 0.87(0.

Thickness 0.31(0.06-1.39) 0.131 0.31(0.

B. MULTIVARIABLE MODEL

Age 1.03(0.99-1.07) 0.199 1.03(0.

Sex, % male 0.88(0.48-1.63) 0.693 0.88(0.
APOE ε4+, % 1.94(1.02-3.76) 0.047 1.94(1.

MMSE 0.97(0.86-1.10) 0.634 0.97(0.

Thickness 0.41(0.07-2.42 0.331 0.41(0.

Table 3. Factors contributing to T(au) status in non-A

normal individuals

Non-AD neurodegen

Braak I-II
(T+ = 1, T- = 0)

OR (95% CI) P OR (

A. UNIVARIATE MODEL

Age 0.99(0.96-1.02) 0.481 1.09(

Sex, % male 1.38(0.81-2.42) 0.246 1.01(
APOE ε4+, % 0.97(0.48-1.89) 0.935 2.55(
Aβ+, % 1.64(0.90-2.92) 0.099 18.3(

MMSE 0.95(0.91-0.99) 0.026 0.89(

Thickness 0.19(0.05-0.68) 0.011 0.03(

B. MULTIVARIABLE MODEL

Age 0.95(0.91-0.97) 0.039 1.03(

Sex, % male 2.08(1.01-4.50) 0.051 1.13(
APOE ε4+, % 0.68(0.30-1.47) 0.339 1.36(
Aβ+, % 3.14(1.43-7.01) 0.004 17.5(

MMSE 0.93(0.87-0.99) 0.025 0.91(

Thickness 0.14(0.03-0.78) 0.024 0.24(

Cognitively norm

OR (95% CI) P OR (

A. UNIVARIATE MODEL

Age 1.06(1.03-1.09) <0.001 1.06(

Sex, % male 0.67(0.35-1.125) 0.220 0.55(
APOE ε4+, % 2.26(1.20-4.33) 0.012 4.81(
Aβ+, % 5.87(3.10-11.6) <0.001 14.80

MMSE 0.87(0.71-1.08) 0.189 0.76(

Thickness 0.01(0.00-0.02) <0.001 0.01(

B. MULTIVARIABLE MODEL

Age 1.01(0.98-1.05) 0.520 1.00(

Sex, % male 0.66(0.32-1.33) 0.258 0.44(
APOE ε4+, % 1.53(0.74-3.21) 0.256 3.10(
Aβ+, % 4.53(2.12-10.0) <0.001 9.71(

MMSE 0.98(0.77-1.28) 0.891 0.74(

Thickness 0.02(0.00-1.34) <0.001 0.03(

Conclusion: We identified several demographic, clini
explain the variance seen in tau PET retention in MCI

Keywords: Tau status, flortaucipir, RO948, AD, non-A

43

AD dementia and in Aβ+ Mild cognitive impairment

ementia

Braak I-IV Braak V-VI

(T+ = 1, T- = 0) (T+ = 1, T- = 0)

5% CI) P OR (95% CI) P

.85-0.95) <0.001 0.88(0.85-0.95) <0.001
.43-1.93) 0.789 0.71(0.43-1.18) 0.188
.14-5.97) 0.025 1.24(0.72-2.12) 0.430
.72-0.89) <0.001 0.90(0.85-0.95) <0.001
.16-2.73) 0.577 0.12(0.02-0.54) 0.007

.82-0.94) <0.001 0.87(0.83-0.91) <0.001
.53-3.40) 0.553 0.69(0.37-1.26) 0.227
.00-6.10) 0.051 1.20(0.63-2.24) 0.579
.68-0.91) 0.001 0.92(0.85-0.99) 0.033
.00-3.43) 0.266 0.05(0.00-0.33) 0.003
ve impairment
5% CI) P OR (95% CI) P

.97-1.04) 0.770 0.95(0.92-0.99) 0.013
.38-1.20) 0.178 0.79(0.41-1.51) 0.470
.99-3.56) 0.054 1.81(0.86-4.05) 0.132
.78-0.98) 0.025 0.91(0.80-1.04) 0.153
.07-1.39) 0.131 0.02(0.01-0.03) <0.001

.99-1.07) 0.051 0.88(0.83-0.93) <0.001
.48-1.63) 0.248 0.79(0.34-1.78) 0.571
.02-3.76) 0.002 4.83(1.84-10.5) 0.002
.86-1.10) 0.021 0.87(0.74-1.01) 0.077
.072.42) <0.001 0.09(0.00-0.38) <0.001

AD neurodegenerative disorders and in cognitively

nerative disorders

Braak I-IV Braak V-VI

(T+ = 1, T- = 0) (T+ = 1, T- = 0)

(95% CI) P OR (95% CI) P

(1.04-1.14) <0.001 1.07(1.00-1.14) 0.040
(0.51-2.06) 0.977 1.47(0.56-4.30) 0.453
(1.19-5.50) 0.016 1.92(0.65-5.55) 0.222
(7.34-55.6) <0.001 21.8(5.99-140.5) <0.001
(0.85-0.94) <0.001 0.91(0.85-0.98) 0.006
(0.01-0.05) <0.001 0.03(0.00-0.06) <0.001

(0.96-1.11) 0.413 0.95(0.86-1.04) 0.274
(0.43-3.07) 0.813 1.28(0.36-4.93) 0.705
(0.51-3.64) 0.542 0.75(0.20-2.68) 0.658
(5.34-80.2) <0.001 36.3(5.80-700.3) 0.001
(0.83-0.98) 0.018 0.95(0.86-1.06) 0.383
(0.02-2.60) 0.235 0.06(0.00-1.10) 0.059
mal individuals
(95% CI) P OR (95% CI) P

(1.02-1.12) 0.009 1.05(0.99-1.12) 0.117
(0.19-1.39) 0.225 0.55(0.12-2.02) 0.396
(1.83-15.0) 0.003 6.23(1.54-41.2) 0.022
0(4.86-64.2) <0.001 9.75(2.41-65.2) 0.004
(0.58-1.02) 0.051 0.63(0.44-0.92) 0.011
(0.00-0.02) <0.001 0.03(0.02-0.04) 0.001

(0.93-1.06) 0.938 0.98(0.54-1.08) 0.551
(0.13-1.25) 0.141 0.44(0.08-1.91) 0.307
(1.03-10.8) 0.055 4.20(0.88-3.10) 0.098
(2.75-47.1) 0.001 7.14(1.33-5.94) 0.017
(0.53-1.06) 0.088 0.58(0.37-0.92) 0.036
(0.00-0.09) 0.002 0.02(0.00-0.14) 0.007

ical and neurobiological factors that are important to
I and AD.

AD

32

Ozlen, Hazal

P154: Widespread amyloid is necessar
entorhinal cortex and cognitive declin

Hazal Ozlen1,2, Alexa Pichet Binette1,2,4, Theres
Villeneuve1,2,3,4

1Douglas Mental Health University Institute, Centre f
(StoP-AD), Montréal, QC, Canada
2Department of Psychiatry, McGill University, Montré
3Department of Neurology and Neurosurgery, McGill
4McGill Centre for Integrative Neuroscience, McGill
Objective: The amount and the spatial distribution of
signals are unknown. Given increasing evidence sugg
adults is biologically relevant, we sought to assess the
increased tau-PET signal and cognitive decline in asym

Methods: One hundred and twenty-nine cognitively u
(PREVENT-AD cohort) underwent Aβ ([18F]NAV469
Gaussian-mixture models to create region-specific thr
previously to be sensitive to early Aβ accumulation (V
Aβ-positive in all regions were classified as the Wides
more regions were included in the Regional Aβ group
compared the three groups’ demographics, tau-PET bi
IV, Ossenkoppele et al., JAMA 2018), subjective mem
Linear mixed-effects models also compared cognitive
(Table 1).
Results: The Regional Aβ group had elevated tau-PET
temporal gyrus when compared with the Aβ-negative
elevated tau PET signal compared with the two other
individuals with Widespread Aβ (and associated tau) s
other two groups did not decline or differed at baselin

Conclusions: Elevated tau-PET signal outside of the e
are detected when Aβ deposition is widespread across

43

ry to detect tau-PET signal beyond the
ne

sa Köbe1,2, Pierre-Francois Meyer1,2,4, Sylvia

for Studies on the Prevention of Alzheimer’s Disease
éal, QC, Canada
l University, Montréal, QC, Canada
University, Montréal, QC, Canada
f Aβ-pathology necessary to detect elevated tau-PET
gesting that subthreshold Aβ accumulation in older
e amount and spread of Aβ burden needed to measure
mptomatic individuals at risk of AD.
unimpaired individuals with a family history of AD
94) and tau ([18F]AV1451) PET scans. We used
resholds of Aβ positivity in seven regions identified
Villeneuve et al, Brain, 2015). Individuals who were
spread Aβ group; those who were positive in one or
p, while the others were considered Aβ-negative. We
inding (in five regions corresponding to Braak I to
mory complaint and baseline cognitive performance.
e trajectories over up to seven years of assessments

T binding in the entorhinal cortex and middle
group (Fig.1). The Widespread Aβ group had
groups across all five regions investigated. Only
showed a faster cognitive decline over time, while the
ne (Fig.2).
entorhinal cortex and measurable cognitive decline
s the cortex.

33

43

34

Keywords: Preclinical AD, amyloid, tau, PET, Cognit
43

tive decline
35

Provost, Karine

P155: Using 18F-Flortaucipir visual as
AT(N) framework: evaluation of intra

Karine Provost1, Leonardo Iaccarino1, David So
Renaud La Joie1, Lauren Edwards1, Amelia Stro
Janabi2, Suzanne Baker2, William Jagust2,3, Gil

1Memory and Aging Center, Department of Neurology
Francisco, CA, US
2Lawrence Berkeley National Laboratory, Berkeley, C
3Helen Wills Neuroscience Institute, University of Cal
4Department of Radiology and Biomedical Imaging, U
Francisco, CA, US

Objectives: To evaluate intra- and inter-rater reliabili
to define T-status, according to the AT(N) research fra
with clinical diagnoses of mild cognitive impairment
AD dementia.

Methods: We included participants from two cohorts
clinical diagnoses of MCI (n= 95, 53 Aß+) , AD deme
Aß+), and cognitively normal controls (n= 80, 28 Aß+
minutes post injection. Amyloid-PET was used to def
reviewed independently by 2 raters who were blinded
imaging and 18F-Flortaucipir SUVR values, using the
(normal, mild temporal binding, AD-like pattern and n
(figure 1). 20 randomly selected scans from each coho
reliability were assessed for binary T-status with Cohe

Results: Intra-rater reliability ranged from substantial
κ(2)=0.76 (95%CI 0.57-0.95); UCSF κ(1)=1.00 (95%
κ(1)=0.70 (95%CI 0.40-1.00), κ(2)=0.66 (95%CI 0.31
(κ=0.68, 95%CI 0.59-0.76), and was higher for the UC
ADNI cohort (κ=0.59, 95%CI 0.48-0.71). Agreement
AD dementia and MCI, and lowest for non-AD demen
patients than for amyloid negative patients (Figure 2).

Conclusions: Our results show that visual assessment
define T-status in the AT(N) research framework, esp
MCI.

43

ssessment to define T-status in the
a- and inter-rater reliability

oleimani-Meigooni1, Orit Lesman-Segev1,
om1, Julie Pham1, Taylor Mellinger1, Mustafa
D. Rabinovici1,2,3,4

y, University of California San Francisco, San

CA, US
lifornia Berkeley, Berkeley, CA, US
University of California San Francisco, San

ity of visual assessment of 18F-Flortaucipir Tau PET
amework, in a cohort of healthy controls and patients
(MCI), Alzheimer’s disease (AD) dementia, or non-

s (UCSF, n= 99 and ADNI n=180, Table 1) with
entia (n=66, 63 Aß+), non-AD dementia (n=38, 5
+), who underwent 18F-Flortaucipir-PET 80-100
fine amyloid status. 18F-Flortaucipir-PET images were
d to clinical diagnosis, amyloid status, structural
NIH color scale. Visual assessment of pattern
non-AD like pattern) was used to attribute T-status
ort were read twice by each rater. Intra- and inter-rater
en’s kappa(κ).
l to perfect (overall κ(1)=0.83 (95%CI 0.67-0.99),
%CI 1.00-1.00), κ(2)=0.90 (95%CI 0.70-1.00); ADNI
1-1.00). Overall inter-rater reliability was substantial
CSF cohort (κ=0.82, 95%CI 0.71-0.94) than for the
was highest for patients with clinical diagnoses of
ntia. Agreement was also higher for amyloid-positive
.

t of 18F-Flortaucipir PET is a reliable method to
pecially in patients with clinical diagnoses of AD and

36

43

37

43

38

Keywords: Tau PET, Tau imaging, Flortaucipir, Alzh
43

heimer’s disease, biomarkers
39

Provost, Karine

P156: Defining T-status in the AT(N) f
Flortaucipir visual assessment, SUVR

Karine Provost1, Leonardo Iaccarino1, David So
Renaud La Joie1, Niklas Mattsson2, Oskar Hans
Amelia Strom1, Julie Pham1, Taylor Mellinger1,
Jagust4,5, Gil Rabinovici1,4,5,6

1Memory and Aging Center, Department of Neurology
Francisco, CA, US
2Clinical Memory Research Unit, Lund University, Lu
3Roche Diagnostics GmbH, Penzberg, Germany
4Lawrence Berkeley National Laboratory, Berkeley, C
5Helen Wills Neuroscience Institute, University of Cal
6Department of Radiology and Biomedical Imaging, U
Francisco, CA, US

Objectives: To compare T-status derived from 18F-Fl
SUVR quantification, and CSF pTau.

Methods: We included participants from two cohorts
diagnoses of MCI (n=94, 53 Aß+) , Alzheimer’s disea
(n=38, 5 Aß+), and cognitively normal controls (n= 80
100 minutes post injection) and CSF analysis within 1
status. T-status was derived from 18F-Flortaucipir con
criteria and from a temporal meta-ROI SUVR using p
1.20). T-status for CSF pTau (Roche Elecsys) was der
comparing amyloid-positive AD/MCI vs. amyloid-neg

Results: Overall agreement between the 3 classificati
was assigned in 96%, 92% and 79% of AD patients us
quantification and CSF pTau respectively; vs. 68%, 49
and 54%, 33% and 36% in CN. Using amyloid-positiv
with 18F-Flortaucipir visual assessment, while specific
2). In a subset of patients with autopsy (n=11), T-statu
had an accuracy of 100% for intermediate-to-high AD
64% for CSF pTau. Patients with concordant T-status
higher CSF pTau values, higher 18F-Flortaucipir SUV
positive.

Conclusions: Concordance of T-status derived from C
SUVR quantification is stage dependent. 18F-Flortauc
SUVR thresholds maximize specificity.

44

framework: comparison of 18F-
R quantification and CSF pTau

oleimani-Meigooni1, Orit Lesman-Segev1,
sson2, Udo Eichenlaub3, Lauren Edwards1,
, Mustafa Janabi4, Suzanne Baker4, William

y, University of California San Francisco, San

und, Sweden

CA, US
lifornia Berkeley, Berkeley, CA, US
University of California San Francisco, San

lortaucipir PET visual assessment, 18F-Flortaucipir

s (UCSF n=99, ADNI n=179, Table 1) with clinical
ase (AD) dementia (n=66, 63 Aß+), non-AD disorders
0, 28 Aß+) who underwent 18F-Flortaucipir PET (80-
1 year. Amyloid-PET was used to define amyloid
nsensus blinded visual assessment based on a priori
previously validated thresholds (SUVR 1.27, SUVR
rived using ROC-based thresholds, obtained
gative healthy controls (pTau>22.0, pTau>22.2).

ion criteria ranged from 63% to 95% (Figure 1). T+
sing 18F-Flortaucipir visual assessment, SUVR
9% and 61% in MCI; 11%, 11% and 21% in non-AD;
ve AD/MCI as gold standard, sensitivity was highest
city was highest applying SUVR thresholds (Figure
us derived from 18F-Flortaucipir visual assessment
DNC, compared to 82% for SUVR quantification and
s across modalities were significantly younger, had
VR values, and were more likely to be amyloid

CSF pTau, 18F-Flortaucipir visual assessment and
cipir visual assessment increases sensitivity, while

40

44

41

Keywords: Tau PET, Flortaucipir, Alzheimer’s diseas
44

se, biomarkers, CSF
42

Quiroz, Yakeel T.

P157: Plasma neurofilament light is as
burden in autosomal dominant Alzhei
COLBOS Project

Yakeel T. Quiroz1, Henrik Zetterberg2, Eric Rei
Velez1, Josh Fox-Fuller1, Joseph Arboleda-Vela
Sperling6, Keith Johnson1, Kaj Blennow2, Franc

1Massachusetts General Hospital, Charlestown, MA,
2Clinical Neurochemistry Laboratory, Sahlgrenska Un
Sweden
3Banner Alzheimer’s Institute, Phoenix, AZ, US
4Universidad de Antioquia, Medellin, CO
5Schepens Eye Research Institute, Boston, MA, US
6Brigham and Women’s Hospital, Boston, MA, US

Background: Neurofilament light (NfL) is a promisin
characterized the associations among plasma NfL, age
Colombian kindred with autosomal dominant Alzheim
carriers from non-carriers 22 years before their media
associations among NfL, hippocampal volume, brain
kindred.

Methods: We examined cross-sectional (n = 45, 24 ca
in non-demented PSEN1 mutation carriers and non-ca
tau (18F-Flortaucipir) and amyloid (PiB) PET imagin
measured associations among baseline NfL and hippo
longitudinal change in brain pathology over 2 years.

Results: Compared to non-carriers, carriers had highe
+/- 3.98; non-carriers: 4.79 +/- 2.13 pg/ml), and greate
Higher baseline NfL concentrations were associated w
greater amyloid (r=0.58, p<0.001) and regional tau bu
r=0.35, p=0.01; precuneus: r=0.53, p=0.001), and low
CERAD Word List Recall: r=-0.59, p<0.001). Longitu
greater annual tau burden in inferior temporal lobe (r=
pathology did not remain significant after adjusting by
and cognitive measures remained significant, even aft

Conclusion: These findings provide support of the va
neurodegeneration, pathology and cognitive decline in
They also further support plasma NfL as a useful meth

Keywords: NfL, tau PET, PiB, cognition, autosomal-d

44

ssociated with regional tau tangle
imer's disease: Findings from the

iman3, Justin Sanchez1, Edmarie Guzman-
asquez5, Ana Baena4, Jennifer Gatchel1, Reisa
cisco Lopera4

US
niversity Hospital, Mölndal, Sweden, Mölndal,

ng biomarker of neuronal degeneration. We recently
e and cognitive decline in over 2,000 members of the
mer’s disease, and reported that NfL distinguished
an age at clinical onset of 44. Here we investigated the
pathology and memory in individuals from the same

arriers) and longitudinal (n = 19, 9 carriers) measures
arriers (mean age: 38 years). Participants underwent
ng, MRI, blood sampling and cognitive testing. We
ocampal volume, brain pathology and memory; and

er baseline plasma NfL concentrations (carriers: 9.01
er levels of baseline brain pathology (p<0.001).
with lower hippocampal volume (r=-0.35, p=0.01),
urden (entorhinal: r=0.44, p=0.02; inferior temporal:
wer cognitive scores (e.g. MMSE: r=-0.49, p<0.001;
udinally, higher baseline NfL was associated with
=0.57, p=0.01). Associations between NfL and tau
y baseline amyloid or age. Associations between NfL
ter those adjustments.

alue of NfL as a blood biomarker for AD-related
n individuals at risk to develop Alzheimer’s disease.
hod for tracking disease progression.
dominant Alzheimer’s disease

43

Salinas, Cristian

P158: Application of tau PET as a bio
therapeutic trials: A pharmaceutical i

Cristian Salinas1, Talakad Lohith2, Qi Guo3, Du
Sanjeewa4, Robert Comley3, Cyrille Sur2, Eric H

1Biogen Inc., Cambridge, MA, US
2Merck & Co., Inc., West Point, PA, US
3AbbVie Inc., North Chicago, IL, US
4Cerveau Technologies Inc., Knoxville, TN, US
The success of tau PET as a biomarker of Alzheimer’s
entitles addressing specific objectives by using quanti
practicality and have a strong connection to the releva
Patient selection, stratification (e.g. to predict disease
drug-target engagement and therapeutic effect are crit
making process of therapeutic AD trials. Tau PET is
objectives and alignment within the pharmaceutical in
and Cerveau Technologies started a pre-competitive c
effectively make use of [18F]MK-6240, a second-gene
trials. Specifically, assessment of tau positivity, tau b
investigated. Additionally, formation and harmonizat
contributions from pharma and academia has played a
resource in the ability to explore, develop and apply fi
Herein, we share an initial perspective from a sub-set
feedback to refine our views and align our approaches
that it will produce a significant and valuable impact i
change the course of AD.

44

omarker of Alzheimer’s disease in
industry perspective

ustin Wooten3, Thom Tulip4, Sulantha
Hostetler2, John Beaver1, Laurent Martarello1

s disease (AD) in the context of a therapeutic trial
itative endpoints that can be applied with a sense of
ant biological pathways and mechanism of action.
progression) and response to treatment as it relates to
tical elements supporting the design and decision-
expected to play a fundamental role in these
ndustry is desirable. In 2018, Biogen, Merck, AbbVie
collaboration to explore analytical methodologies that
eration tau PET tracer, in AD clinical
burden (signal intensity), and tau spread are being
tion of a collaborative imaging database with
an important role and provided an indispensable
it-for-purpose methodologies.
of pharma partners on tau PET analysis seeking
s with the wider tau imaging community in the hope
in the development of therapies that are intended to

44

Keywords: Tau PET, Alzheimer’s Disease, MK-6240,
44

, Quantification, Pharmaceutical Industry
45

Sanabria Bohorquez, Sandra

P159: Measuring increases of tau path
[18F]GTP1 (Genentech tau probe 1) P

Sandra Sanabria Bohorquez1, Suzanne Baker1,2,
Jan Marik1, Robby Weimer1

1Genentech, Inc., South San Francisco, CA, US
2Molecular Biophysics and Integrated Bioimaging, La
US
Objective: To quantify tau burden change using [18F
(AD).
Methods: [18F]GTP1 imaging was performed at base
positive cognitive normal subjects (CN; n=2 and 7, re
n=24; MMSE 24-30, CDR = 0.5) and mild/moderate (
subjects. SUVR and Extent were calculated using the
positivity threshold was defined as the average uptake
measurements were performed in the whole cortical g
regions. SUVR and Extent vs. time were assumed to b
was calculated using simple linear regression.
Results: Figures shows baseline measurements, the av
burden change. Tau burden increased globally and reg
In AD, the average GM SUVR and Extent APC (±95%
Prodromal subjects displayed similar increases in SUV
5.7-7.9%, respectively). In MM subjects, no SUVR an
increases of 2.9±2.1% and 3.2±2.1% in the SUVR AP
were observed in Braak 3/4 and 5/6, respectively. Ove
similar for SUVR and Extent.
Conclusion: Similar global tau burden increases were
differences between prodromal and MM subjects. Res
AD tau pathology change in observational and interve

44

hology in Alzheimer's disease using
PET imaging

, Paul Manser1, Balazs Toth1, Edmond Teng1,

awrence Berkeley National Laboratory, Berkeley, CA,

F]GTP1 SUVR and Extent in Alzheimer’s disease

eline, 6, 12 and 18 months in amyloid negative and
espectively), and amyloid positive prodromal (Prod;
(MM; n=30; MMSE 22-30, CDR = 0.5-2) AD

cerebellum gray. For each scan, a voxelwise tau
e in the cerebellum plus 2 standard deviations. Global
gray matter (CG), and regionally, in the in vivo Braak
be linear and the annualized percentage change (APC)

verage APC (±95%CI) and example images of tau
gionally in prodromal and MM groups but not in CN.
%CI) were 2.8±1.2% and 6.2±2.4%, respectively.
VR and Extent across Braak regions (1.9-3.0% and
nd Extent increases were observed in Braak 1/2, but
PR, and of 4.6±3.4% and 6.9±3.8% in the Extent APR
erall, the APC variance relative to the mean rates were

e observed in the AD groups but there were regional
sults support further investigation of Extent to assess
entional studies.

46

44

47

Keywords: Tau PET, [18F]GTP1, longitudinal study,
44

, SUVR, Extent
48

Sanchez, Justin

P160: Evaluation of tau PET staging i

Justin Sanchez1, Alex Becker1, Heidi Jacobs1,3,
Zoe Rubinstein1, Emma Thibault1, Aaron Schul
Rentz1,2, Julie Price1, Reisa Sperling1,2, Keith Jo

1Massachusetts General Hospital, Boston, MA, US
2Brigham and Women’s Hospital, Boston, MA, US
3Maastricht University, Maastricht, The Netherlands
4Université Catholique de Louvain, Brussels, Belgium
5Boston University School of Medicine, Boston, MA, U
6The Framingham Heart Study, Framingham, MA, US
7University of Texas, San Antonio, San Antonio, TX, U
8Grupo de Neurociencias, Universidad de Antioquia,

Introduction: The spatiotemporal progression of cort
proceeds from medial temporal lobe (MTL) to tempor
association with rising levels of amyloid-beta (aβ). W
PET elevation thresholds in a sample of clinically nor
longitudinal PET measures (Sanchez, HAI 2019). Her
validate them in the cross-sectional A4/LEARN data s

Methods: In order to maximize expected between-gro
threshold derivation consisted of 47 high-aβ, cognitiv
unimpaired participants aged 21-49years (Fig.1). App
Flortaucipir (FTP) SUVr data, proxy ROIs defined 3 s
typical area of initial cortical deposition, Tp2, inferior
Tp3, precuneus (PC), representing extra-temporal neo
mean/sd=71.8/4.8) underwent FTP and Florbetapir PE
190CL. We assessed concordance of the staging schem
high-tau participants in each region by age, aβ, and AP

Results: Elevated FTP in A4/LEARN was observed m
(24.7%), followed by PC (7.9%) (Fig.2). 95% of parti
the Tp staging scheme (binomial test vs discordant, p<
associated with greater aβ burden and APOEe4 carrier

Conclusions: Tau PET staging in A4/LEARN using G
accorded with the spatiotemporal progression scheme
generalizability of a regional threshold based Tp stagi
detecting early tau pathology.

44

in the A4/LEARN study

Bernard Hanseeuw1,4, Danielle Mayblyum1,
ltz1, Sudha Seshadri5,6, Yakeel Quiroz1, Dorene
ohnson1,2, on behalf of the A4 Study Team2

m
US
S
US
Antioquia, CO

tical tauopathy, as implied by autopsy studies,
ral neocortex and then to extra-temporal neocortex, in
We previously identified successive, region-specific tau
rmal adults and validated these thresholds with
re we derive thresholds in an independent data set and
set (N=441).

oup difference, the reference sample used for
vely-impaired participants and 71 low-aβ, cognitively-
plying Gaussian mixture models to reference sample
staging thresholds: Tp1, rhinal cortex (RC), the
r temporal (IT), representing temporal neocortex, and
ocortex. A4/LEARN participants (age
ET. Aβ burden range was approximately -20 to
me using binomial tests as well as the proportion of
POE genotype.

most commonly in RC (40.4%), followed by IT
icipants showed FTP tau elevations concordant with
<0.0001, Fig.2). As expected, higher Tp stage was
r frequency (Fig 2,3).

GMM thresholds from an independent sample
e implied by autopsy studies. These results suggest the
ing system, and the utility of MTL ROIs like RC for

49

45

50

45