HAI Book 2020 - Dec 18B

51

Schwarz, Christopher

Tau positivity: comparing flortaucipir
Z-scores

Christopher Schwarz1, Terry Therneau1, Scott P
Matthew Senjem1, Val Lowe1, Kejal Kantarci1,
Jack1

1Mayo Clinic, Rochester, MN, US

Background: Meta-ROI or reporter ROI approaches
biomarker status, but these approaches may not detect
Methods: We computed regional flortaucipir (18F-AV
ADRC participants. Cognitively unimpaired participa
population, and per-region mean (SD) were calculated
We Z-transformed all regional SUVRs and computed
(“ZMax”). We used the 97.5th percentile of ZMax amo
ZMax>5.16) and compared this with our existing tem
parahippocampal, fusiform, inferior temporal, middle
1.25). Using the remaining (n=1267) baseline scans, w
(abnormal meta-ROI) vs. Z+ (abnormal in any ROI).
Results: 457 (36%) scans were Z+, while 442 (35%)
61% were PiB+, and 44% were clinically impaired. T
were PiB+, and 19% were clinically impaired. ZMax
annualized change in auditory verbal learning test (AV
largest signal in a medial-temporal region; 15% in ano
elsewhere. Across all 1267 scans, these were 36%, 30
Discussion: Replacing our meta-ROI approach with Z
individuals as tau-positive. The largest Z-scores typica
whose contributions to meta-ROIs may be weakened w
regions. Our major findings were consistent across sev
with mean-of-top-n-regions, and replacing per-region

35

r meta-ROI vs. maximum of regional

Przybelski1, Heather Wiste1, Jeffrey Gunter1,
Ronald Petersen1, David Knopman1, Clifford

are commonly used to determine abnormal tau
t elevated signal in focal or atypical cortical regions.
V-1451) PET SUVR for Mayo Clinic MCSA and
ants ages 30-50 (n=144) were defined as a normative
d across scans (SD included only SUVRs>median).
the within-person maximum across cortical ROIs
ong the normative group to define abnormality (Z+:
mporal meta-ROI (GlobalTau: entorhinal, amygdala,
temporal) and cut-point (T+: GlobalTau SUVR >=
we compared the populations identified as T+

were T+. T-Z+ individuals were older (median=75y),
T+Z- individuals were younger (median=69y), 35%

was more correlated (Spearman -0.24 vs. -0.16) with
VLT). Among T-Z+ scans (n=80), 54% had the
other (non-medial) temporal region; and 31%
0%, and 34% respectively.
ZMax may capture a more pathologic set of
ally occurred in relatively small temporal lobe regions
when using a mean, rather than a maximum, across
veral variants of ZMax threshold, replacing maximum
median-across-voxels with 95th percentile.

52

35

53

Keywords: flortaucipir,tau,cutoff,positivity,cutpoint
35

54

Friday, January 17, 2020 - 9:15am - 10:00am

P3A: POSTER SESSION

Board Poster Title
#

120 Depression is predicted by tau imaging Babulal | Roe | Sto
biomarker among cognitively normal Morris | Ances
adults

Temporal evaluation for maximum Wisch | Gordon | S
172 CSF pTau181 – [18F]Flortaucipir Benzinger | Morris

concordance

Cerebrospinal fluid tau phosphorylated

at amino acid 181 or 231: a biomarker Benedet | J. Ashton

144 comparison in the detection of Vanmechelen | K.

neurofibrillary tangle pathology in the Therriault | Chamo

AD spectrum

Local associations between tau and Bilgel | Wong | Re
121 neurodegeneration among non-

demented older adults

Brendel | Beyer | N

Marek | Song | Pal

Perfusion-phase 18F-PI-2620 tau-PET | Barbe | Onur | Jes

122 imaging as a surrogate marker of Rullmann | Schilda

neuronal injury Russel | Stephens |

Classen | Hoegling

| Seibyl | Sabri

Brendel | Song | Be

123 Binding characteristics of 18F-PI-2620 Palleis | Kaiser | R
differentiate the clinically predicted tau Barbe | Onur | Jess
isoform in suspected 3/4-repeat and 4- Rullmann | Schilda
repeat tauopathies Russel | Stephens |
Classen | Hoegling

| Seibyl | Sabri

Brendel | Palleis | P

18F-PI2620 tau-PET for assessment of Höllerhage | Sauer

124 heterogeneous neuropathology in Stephens | Barthel

corticobasal syndrome Villemagne | Barte

Hoeglinger

125 Optimal reference region for the Bullich | Müller | R
quantification of tau load in the brain Santi | Koglin | Ste
using 18F-PI-2620 PET

126 Quantifying tau PET imaging reliably Chen | Gordon | M
in the presence of off-target binding Flores | Keefe | Pau
Benzinger

127 [18F]THK5317 imaging as a predictive Chiotis | Savitchev
tool of prospective cognitive decline in Antoni | Nordberg
prodromal and dementia-stage AD

128 Comparison of quantitative cutoff Minhas | Laymon |
methods for [18F]AV-1451 Aizenstein | Lopez

129 Pilot study on the relationship between Fernandez Arias | P
18F-MK-6240 and VBM in early and | Savard | Ottoy | M
late stages of AD Thomas | Stevenso
Soucy | Gauthier |

Functional connectivity brain Franzmeier | Neitz
130 architecture predicts the rate of tau Ossenkoppele | Ha

accumulation in AD

How innocent is PART?: Mesial

131 temporal tau is associated with worse Groot
cognitive performance in Aβ-negative

cognitively normal individuals

CSF P-tau detects cerebral tau
132 accumulation earlier than tau PET in Guo | Jagust | Land

amyloid positive elderly adults

35

N 3A

Authors Presenter Page

out | Rajasekar | Wisch | Benzinger | Ances, Beau 359

Schindler | Flores | Dincer | Fagan | Ances, Beau 479
s | Ances

n | A. Pascoal | Stoops | Francois | 412
Karikari | Mathotaarachchi | Savard | Benedet, Andrea
oun | Zetterberg | Blennow | Rosa-Neto

esnick Bilgel, Murat 360

Nitschmann | Barthel | van Eimeren | 363
lleis | Respondek | Sauerbeck | Hammes
ssen | Saur | Schroeter | Rumpf |
an | Patt | Neumaier | Barret | Madonia | Brendel, Matthias
| Roeber | Herms | Bötzel | Levin |
ger | Bartenstein | Villemagne | Drzezga

eyer | Barthel | van Eimeren | Marek | 365
Respondek | Sauerbeck | Hammes |
sen | Saur | Schroeter | Rumpf |

an | Patt | Neumaier | Barret | Madonia | Brendel, Matthias
| Roeber | Herms | Bötzel | Levin |
ger | Bartenstein | Villemagne | Drzezga

Prix | Gehmeyr | Bötzel | Danek | Brendel, Matthias 367
rbeck | Beyer | Nitschmann | Song |
| Patt | Sabri | Drzezga | van Eimeren |
enstein | Perneczky | Haass | Levin |

Roé-Vellvé | Jovalekic | Perrotin | De Bullich Roig, 369
370
ephens Santiago

McCullough | Zaza | Mejias | Dincer |
ulick | Jackson | Koudelis | Su | Morris | Chen, Charles

va | Poulakis | Saint-Aubert | Wall | Chiotis, 372
g Konstantinos 374
375
| Lopresti | Snitz | Tudorascu | Cohen, Ann 377
z | Mathis | Klunk | Cohen

Pascoal | Benedet | Thierrault | Su Kang

Mathotaarachchi | Lussier | Tissot | Fernandez Arias,

on | Rahmouni | Wang | Massarweh | Jaime

Rosa-Neto

zel | Rubinski | Smith | Strandberg | Franzmeier,
ansson | Ewers Nicolai

Groot, Colin 381

dau Guo, Tengfei 384

55

Board Poster Title
#

133 Amyloid, tau, and atrophy in Hanseeuw | Jacobs
preclinical AD: A longitudinal study Scott | Hampton | S

Increased levels and phosphorylation Hansson | Anderss

134 of soluble tau proteins occur earlier Strandberg | Zetter

than changes in Tau PET in AD | Smith | Palmqvist

Preclinical characterization of Harada | Lerdsirisu
135 [18F]THK-5562, a novel tau PET Arai | Yanai | Kudo

tracer with little off-target binding

Comparison of tau PET imaging using

136 18F-APN-1607 and 18F-THK5351 in Hsiao | Lin | Huang
Alzheimer disease patients and normal

controls

Iaccarino | La Joie

137 In vivo amyloid-PET and tau-PET | Provost | Collins
evidence in early-onset AD: taking the Foroud | Gatsonis
LEADS Vemuri | Day | Gra
Mendez | Onyike |

Koeppe | Dickerso

138 Multimodal in vivo investigation of Iaccarino | La Joie
amyloid and tau biomarkers Pham | Chaudhary
associations with cerebrospinal fluid Miller | Jagust | Fa
NfL and YKL40 levels

139 Activated microglia and amyloid load Kang | Ottoy | Sav
potentiate tau deposition leading to Benedet | Chamou
cognitive dysfunction in AD | Stevenson | Rahm
Soucy | Gauthier |

Tauopathy in females is more Kang | Ottoy | Cha
140 vulnerable to amyloid or Benedet | Pascoal |
Stevenson | Rahmo
neuroinflammation in AD Massarweh | Soucy

Prediction of brain tau accumulation in

141 amyloid positive cognitive impairment Kim | Park | Park |
patients using multimodal biomarkers Seo

with machine learning approach

Comparison of longitudinal change

142 metrics for 18F-RO948 PET among Klein | Sanabria | L
cognitively unimpaired and patients Stomrud | Smith | H
with MCI or AD dementia in the

BioFINDER2 study

Diagnostic performance of

[18F]RO948 tau positron emission Leuzy | Smith | Os

145 tomography in the differentiation of Olsson | Jögi | Palm

AD from other neurodegenerative Stomrud | Hansson

disorders

Longitudinal changes in tau pathology
measured by [18F]RO948 tau-PET are Leuzy | Klein | Oss
146 associated with elevated CSF P-tau: Palmqvist | Strand
Preliminary findings from the Swedish Smith | Hansson
BioFINDER-2 study

A comparison of ischemic stroke-

147 induced changes on 18F-APN-1607 Lin | Hsiao | Huang
(18F-PMPBB3) and 18F-THK-5351

uptake patterns

Longitudinal change in [18F]GTP1

148 SUVR over 18 months depends on Manser | Sanabria
baseline SUVR intensity and spatial Marik | Weimer
distribution and shows trends with

cognitive decline

149 18F-APN-1607: a promising PET Margolin | Lin | Te
tracer for multiple tauopathies Sandiego | Huang
Lu | Zuo | Higuchi

150 Entorhinal cortical tau accumulation is Mecca | Chen | Na
inversely associated with hippocampal Bartlett | Zhao | Ga
synaptic density in older individuals Carson | van Dyck
with normal cognition and early AD

35

Authors Presenter Page

s | Schultz | Buckley | Properzi | Farrell | Hanseeuw, 388
391
Sanchez | Sperling | Johnson Bernard
392
son | Janelidze | Ossenkoppele | Insel |
rberg | Blennow | Chai | Dage | Stomrud Hansson, Oskar
t | Mattsson

uk | Yiqing | Ezura | Shimizu | Morito | Harada, Ryuichi
o | Furumoto | Okamura

g | Huang Hsiao, Ing-Tsung 393

e | Lesman-Segev | Soleimani-Meigooni 395
| Aisen | Borowski | Eloyan | Fagan | 399
| Jack Jr. | Kramer | Saykin | Toga | Iaccarino, 403
aff-Radford | Honig | Jones | Masdeu | Leonardo 404
| Rogalski | Salloway | Wolk | Wingo | 406
on | Apostolova | Carrillo | Rabinovici

e | Edwards | Lesman-Segev | Strom | Iaccarino,
y | Fenton | Jerome | Janabi | Baker | Leonardo
agan | Rabinovici

vard | Pascoal | Mathotaarachchi |

un | Therriault | Lussier | Tissot | Thomas Kang, Min Su
mouni | Arias | Wang | Massarweh |

Rosa-Neto

amoun | Mathotaarachchi | Savard | Kang, Min Su
| Therriault | Lussier | Tissot | Thomas |
ouni | Fernandez-Arias | Wang |
y | Gauthier | Rosa-Neto

Kang | Kim | Jang | Kim | Na | Lee | Kim, Jaeho

Leuzy | Borroni | Mattsson | Palmqvist | Klein, Gregory 408
Hansson

ssenkoppele | Santillo | Borroni | Klein |

mqvist | Mattsson | Strandberg | Leuzy, Antoine 415

n

senkoppele | Mattsson | Janelidze | Leuzy, Antoine 418
dberg | Coloma | Borroni | Stomrud |

g | Huang Lin, Kun-Ju 420

Bohorquez | Teng | Baker | Toth | Manser, Paul 421

empest | Chen | Marek | Russell | Margolin, Richard 423
| Hsiao | Tamagnan | Alagille | Guan | 425
i | Jang

aganawa | Toyonaga | Godek | Harris | Mecca, Adam
allezot | Nabulsi | Huang | Arnsten |
k

56

Board Poster Title
#

Mental and physical activity during the

151 uptake period affects off-target binding Min | Apgar | Scot
in extra- and within-brain Tau PET Botha | Vemuri | G

(18F-FTP)

Validation of clinical protocols for Koran | Shams | Ad

143 clinicians analyzing 18F-PI-2620 tau | Castillo | Hall | Sh

PET/MRI images Zaharchuk | David

152 In vivo uptake of 18F-PM-PBB3 (18F- Nakano | Shimada
APN-1607) in patients with | Takado | Shinotoh
corticobasal syndrome Ono | Tempest | Ja
Kuwabara | Higuch

153 Factors predicting tau PET status in Ossenkoppele | Le
cognitively unimpaired and impaired Matsson | Palmqvi
individuals

Widespread amyloid is necessary to

154 detect tau-PET signal beyond the Ozlen | Binette | K

entorhinal crtex and cognitive decline

155 Using 18F-Flortaucipir visual Provost | Iaccarino
assessment to define T-status in the | La Joie | Edwards
AT(N) framework: evaluation of intra- Baker | Jagust | Ra
and inter-rater reliability

Defining T-status in the AT(N) Provost | Iaccarino

156 framework: comparison of 18F- | La Joie | Mattsson

Flortaucipir visual assessment, SUVR Strom | Pham | Me

quantification and CSF pTau Rabinovici

157 Plasma neurofilament light is Quiroz | Zetterberg
associated with regional tau tangle Fox-Fuller | Arbol
burden in autosomal dominant AD: Sperling | Johnson
Findings from the COLBOS Project

158 Application of tau PET as a biomarker Salinas | Lohith | G
of AD in therapeutic trials: A Comley | Sur | Hos
pharmaceutical industry perspective

159 Measuring increases of tau pathology Sanabria Bohorque
in AD using [18F]GTP1 (Genentech Marik | Weimer
tau probe 1) PET imaging

160 Evaluation of tau PET staging in the Sanchez | Becker |
A4/LEARN study Rubinstein | Thiba
Price | Sperling | Jo

Clinical evaluation o[18F]-JNJ- Schmidt | Janssens

161 64326067, a candidate PET tracer for Barret | Constantin

the detection of tau pathology in AD Kolb

Association between cerebrospinal

fluid neurofilament light chain and

162 markers of neurofibrillary Schultz | Schindler
pathophysiology: Findings from the Gordon | Benzinge

Knight Alzheimer Disease Research

Center

18F-PM-PBB3 (18F-APN-1607) Shimada | Tagai | T

163 uptake associates with plasma NfL Takado | Shinotoh
level and motor disability in patients Ono | Hirano | Tem

with progressive supranuclear palsy Zhang | Tokuda | H

The rate of accumulation of tau Smith | Mattsson |
164 aggregates is higher in females and Ossenkoppele | Ha

younger individuals

165 Resting state functional connectivity Svaldi | Goñi | Stag
associations with F18-Florbetapir PET Risacher | Saykin |
versus F18-Flortaucipir PET

166 Tau positivity: Comparing flortaucipir Therneau | Schwar
meta-ROI vs maximal single region Kantarci | Petersen

Intrinsic connectivity of the human Therriault | Pascoa
167 brain provides scaffold for tau Chamoun | Gauthi
Rosa-Neto
aggregation in AD

168 Changes in volumetric MRI measures Toth | Sanabria Bo
at 12 months significantly correlate Marik | Weimer

35

Authors Presenter Page

tt | Hol | Lundt | Albertson | Schwarz | Min, Hoon-Ki 427
Gunter | Petersen | Jack | Lowe

dams | Cazevedo | Toueg | Corso | Hunt Mormino, 410
ha | Fredericks | Greicius | Wagner | Elizabeth 430
dzon | Chin | Mormino 431

a | Tagai | Matsuoka | Kubota | Takahata

h | Yamamoto | Sano | Seki | Hirano | Nakano,

ang | Sahara | Kawamura | Zhang | Yoshikazu

hi

euzy | Cho | La Joie | Strandberg | Ossenkoppele, Rik
ist | Lyoo | Rabinovici | Smith | Hansson

Köbe | Meyer | Villeneuve Ozlen, Hazal 433

o | Soleimani-Meigooni | Lesman-Segev 436
s | Strom | Pham | Mellinger | Janabi | Provost, Karine 440
abinovici

o | Soleimani-Meigooni | Lesman-Segev

n | Hansson | Eichenlaub | Edwards | Provost, Karine
ellinger | Janabi | Baker | Jagust |

g | Reiman | Sanchez | Guzman-Velez |

leda-Velasquez | Baena | Gatchel | Quiroz, Yakeel 443

n | Blennow | Lopera

Guo | Wooten | Tulip | Sanjeewa | Salinas, Cristian 444
stetler | Beaver | Martarello

ez | Baker | Manser | Toth | Teng | Sanabria 446
Bohorquez, Sandra

| Jacobs | Hanseeuw | Mayblyum | 449
ault | Schultz | Seshadri | Quiroz | Rentz | Sanchez, Justin 453
ohnson | the A4 Study Team

s | Moechars | Rombouts | Timmers |
nescu | Madonia | Russell | Sandiego | Schmidt, Mark

r | Chen | Sutphen | Morris | Fagan | Schultz, Stephanie 454
er

Tatebe | Matsuoka | Kubota | Takahata |

h | Yamamoto | Sano | Seki | Nakano | Shimada, Hitoshi 457
mpest | Jang | Sahara | Kawamura | 458

Higuchi

Pontecorvo | Devous | Strandberg | Smith, Ruben
ansson

ge | Abbas | Dzemidzic | West | Svaldi, Diana 462
| Apostolova 466
469
rz | Wiste | Gunter | Senjem | Lowe | Therneau, Terry
n | Knopman | Mielke | Jones | Jack

al | Savard | Mathotaarachchi | Benedet |

ier | Saha-Chaudhuri | Massarweh | Therriault, Joseph

ohorquez | Manser | Baker | Teng | Toth, Balasz 473

57

Board Poster Title
#

with baseline [18F]GTP1 SUVR, but
not [18F]GTP1 change

Elevated medial temporal lobe Tau Toueg | Deutsch | C
Harrison | Azevedo
169 PET with 18F-PI2620 in normal Sha | Davidzon | C
controls with “borderline”

neuropsychological testing profiles

Hippocampal tau accumulation Trelle | Toueg | Ca

170 predicts individual differences in Jayakumar | Nadia

episodic memory in cognitively normal Davidson | Deutsc

older adults Carr | Chin | Wagn

Temporo-limbic tau burden, as

171 measured by 18F-AV1451, is a Fischer | Ahnen | D

correlate of hippocampal volume loss Kagerer | Gietl | H

over time

[18F]flortaucipir PET is more closely Wolters | Ossenko

173 associated with disease severity than Visser | Tuncel | G

CSF p-tau in AD Flier | Teunissen |

Associations of tau pathology and

174 functional connectivity with Shafer | Bilgel | Zi
retrospective cognitive change among

cognitively normal older adults

35

Authors Presenter Page

Castillo | Hunt | Corso | Trelle | Toueg, Tyler 475
o | Shen | Anders | Hall | Fredericks | Trelle, Alexandra 476
Chin | Khalighi | Wagner | Mormino 477
481
astillo | Channappa | Corso | Hunt |
adwala | Guo | Azevedo | Shin |
ch | Hall | Sha | Fredericks | Kerchner |
ner | Mormino

Dissertori | Lustermann | Weber | Buck | Unschuld, Paul
Hock | Unschuld

oppele | Verfaillie | Coomans | Timmers |
Golla | Windhorst | Boellaard | van der Wolters, Emma

Scheltens | van Berckel

iontz | Wong | Resnick Ziontz, Jacob 485

58

Ances, Beau

P120: Depression is predicted by tau i
normal adults

Ganesh Babulal1, Catherine Roe1, Sarah Stout1,
Benzinger1, John Morris1, Beau Ances1

1Washington University School of Medicine, St. Louis
Objective: Depression is both a risk factor and conseq
studies using cerebrospinal fluid and amyloid Positron
levels of AD biomarkers are associated with the devel
study examines whether tau and amyloid imaging, pre
normal adults. Secondary analysis assesses how antid
Methods: Cognitively normal participants (Clinical D
Alzheimer’s Disease Research Center at Washington
whether, in vivo tau PET (Flortaucipir), predicted dep
including age, sex, education, race, and apolipoprotein
interaction between tau PET and antidepressant use in
covariates. A second set of models were conducted su
(Florbetapir). We also examined group differences on
and a neuropsychological composite score.
Results: Data were available from 301 cognitively no
69.5 (SD=8.0)). Participants with elevated tau PET we
between tau and antidepressant use was also significan
antidepressants had a greater risk of depression. Relat
predicting depression. There were no differences betw
neuropsychiatric symptoms, and a neuropsychologica
Conclusions: Our results demonstrate that tau, not am
there was no relationship with depressive symptoms. A
increase the risk of depression among cognitively norm
Keywords: Tau PET, Amyloid PET, Depression, Antid

35

imaging biomarker among cognitively

Ganesh Rajasekar1, Julie Wisch1, Tammie

s, MO, US
quence of Alzheimer’s disease (AD). Biomarker
n Emission Tomography (PET) have found elevated
lopment of depressive symptoms over time. This
edicts a depression diagnosis among cognitively
depressant modifies this relationship.
Dementia Rating = 0) were enrolled from the Knight
University. Logistic regression models evaluated
pression diagnosis, after adjusting for covariates
n ε4. Similarly, a second model tested the statistical
n predicting depression, while adjusting for the same
ubstituting tau PET with amyloid PET imaging
n depressive symptoms, neuropsychiatric symptoms,

ormal participants (46.2 to 91.3 years old and mean of
ere twice more likely to be depressed. The interaction
nt suggesting those with elevated tau and taking
tedly, amyloid was not statistically significant in
ween groups on depressive symptoms,
al composite score.
myloid PET predicts a depression diagnosis, however,
Additionally, antidepressant use interacts with tau to
mal adults.
depressants

59

Bilgel, Murat

P121: Local associations between tau a
demented older adults

Murat Bilgel1, Dean Wong2, Susan Resnick1

1Laboratory of Behavioral Neuroscience, National Ins
2Department of Radiology and Radiological Science,
Baltimore, MD, US
Introduction: We propose a framework for studying
neurodegeneration, as measured by cerebral blood flo
Methods: We used cross-sectional data for 50 non-de
Study of Aging. For each participant, we quantified th
(PiB-R1) within the cerebral cortex using geographica
separately assessed the local associations of tau and G
were adjusted for local amyloid burden (PiB-DVR). U
participant, we performed a t-test at each voxel of the

36

and neurodegeneration among non-

stitute on Aging, Baltimore, MD, US
Johns Hopkins University School of Medicine,
individual-level local associations between tau and
ow (CBF) and gray matter (GM) volume.
emented participants from the Baltimore Longitudinal
he local associations of tau (FTP-SUVR) with CBF
ally weighted regressions (Figure 1). We also
GM volume (measured by RAVENS). All models
Using the resulting association maps for each
tau-CBF and tau-GM association maps.

60

Results: Higher FTP-SUVR was associated with high
36

her PiB-DVR throughout the cerebral cortex. These
61

widespread associations remained in analyses stratifie
Adjusting for PiB-DVR, higher FTP-SUVR was assoc
frontal gyrus, inferior and middle temporal gyri, fusifo
gyrus. There were negative associations in the left cen

Adjusting for local PiB-DVR, higher FTP-SUVR was
planum temporale, transverse and superior temporal g
were positive associations in the right inferior occipita

Discussion: The widespread association between FTP
suggests that there is a common source of signal in the
between FTP-SUVR and PiB-R1 support the notion of
accumulation. The negative associations between FTP
associated with greater insular atrophy in non-dement
This research was supported by the Intramural Resear
Keywords: FTP, PiB, CBF, neurodegeneration, tau

36

ed by PiB status.
ciated with higher PiB-R1 bilaterally in the superior
orm, posterior cingulate, precuneus, and lingual
ntral operculum and insula (Figure 2).

s associated with lower GM volume in the right
gyri, and bilaterally in the insula and operculum. There
al gyrus (Figure 3).

P-SUVR and PiB-DVR, regardless of PiB status,
e FTP and PiB scans. The positive associations
f a period of hyperactivity associated with initial tau
P-SUVR and GM volume suggest that greater tau is
ted individuals.
rch Program of the NIH, National Institute on Aging.

62

Brendel, Matthias

P122: Perfusion-phase 18F-PI-2620 ta
marker of neuronal injury

Matthias Brendel1, Leonie Beyer1, Alexander N
Eimeren3,4,5,6, Ken Marek7,8, Mengmeng Song1,
Sauerbeck1, Jochen Hammes4, Michael Barbe5,
Saur13, Matthias L. Schroeter14,15,16, Jost-Julian
Schildan2, Marianne Patt2, Bernd Neumaier3,4, O
Russel7,8, Andrew Stephens17, Sigrun Roeber18,
Levin6,9, Joseph Classen13, Guenter Hoeglinger6
Villemagne21,22,23, Alexander Drzezga4,6, John S

1Department of Nuclear Medicine, University Hospita
2Department of Nuclear Medicine, University of Leipz
3Cognitive Neuroscience, Institute for Neuroscience a
Juelich, Germany
4Department of Nuclear Medicine, University Hospita
5Department of Neurology, University Hospital Colog
6German Center for Neurodegenerative Diseases (DZ
7InviCRO, LLC, Boston, MA, US
8Molecular Neuroimaging, A Division of inviCRO, Ne
9Department of Neurology, University Hospital of Mu
10Department of Neurology, Technical University Mun
11Center for Memory Disorders, University Hospital C
12Department of Psychiatry, University Hospital Colo
13Department of Neurology, University of Leipzig, Lei
14Clinic for Cognitive Neurology, University of Leipzi
15LIFE - Leipzig Research Center for Civilization Dis
16Max- Planck-Institute of Human Cognitive and Brai
17Life Molecular Imaging GmbH, Berlin, Germany
18Center for Neuropathology and Prion Research, Un
Germany
19Department of Neurology, Medizinische Hochschule
20Munich Cluster for Systems Neurology (SyNergy), M
21Department of Molecular Imaging & Therapy, Austi
22The Florey Institute of Neuroscience and Mental He
Australia
23Department of Medicine, Austin Health, The Univer

Aim: Second generation tau radioligands for use with
developed for visualization of tau deposits in vivo. Fo
radioligands, it has been shown that reduced early-pha
injury. Therefore, we investigated the performance of
2620 for the assessment of neuronal injury.

Methods: Twenty-six subjects were referred with sus
2620 tau-PET (0-60 min p.i.) and 18F-fluorodeoxygluc
uptake-value-ratios of early-phase perfusion images (s
of 18F-PI-2620 were correlated with corresponding qu
normalization). Reduced tracer uptake in cortical targe
dimensional stereotactic surface projections by three e

36

au-PET imaging as a surrogate

Nitschmann1, Henryk Barthel2, Thilo van
Carla Palleis9, Gesine Respondek6,10, Julia
Özgür Onur5, Frank Jessen6,11,12, Dorothee
Rumpf13, Michael Rullmann2, Andreas

Oliver Barret7,8, Jennifer Madonia7,8, David S.
Jochen Herms6,18, Kai Bötzel9, Johannes
6,10,19, Peter Bartenstein1,20, Victor
Seibyl7,8, Osama Sabri2

al of Munich, LMU Munich, Munich, Germany
zig, Leipzig, Germany
and Medicine (INM-3), Research Centre Juelich,

al Cologne, Cologne, Germany
gne, Cologne, Germany
ZNE), Bonn, Germany

ew Haven, CT, US
unich, LMU Munich, Munich, Germany
nich, Munich, Germany
Cologne, Cologne, Germany
ogne, Cologne, Germany
ipzig, Germany
ig, Leipzig, Germany
seases, University of Leipzig, Leipzig, Germany
in Sciences, Leipzig, Germany

niversity Hospital of Munich, LMU Munich, Munich,

e Hannover, Hannover, Germany
Munich, Germany
in Health, Heidelberg, Australia
ealth, The University of Melbourne, Melbourne,

rsity of Melbourne, Melbourne, Australia

h positron emission tomography (PET) have been
or several β-amyloid and first generation tau
ase perfusion can be used as a surrogate of neuronal
f early acquisitions of the novel tau radioligand 18F-PI-

spected tauopathies and received a dynamic 18F-PI-
cose (FDG)-PET (30-50 min p.i.). Regional standard-
single frame SUVr) and the blood flow estimate (R1)
uantification of FDG-PET (global mean/ cerebellar
et regions was interpreted visually using 3-
experienced and three expert readers and intraclass

63

correlation coefficients (ICC) were calculated between
Results: Highest agreement with FDG quantification
2.5 min p.i. for global mean (R > 0.69) and cerebellar
0.5-2.5 min SUVr & R1) displayed excellent agreemen
0.75 ± 0.10, RR1 = 0.75 ± 0.10) and cerebellar normali
interpretation revealed high regional ICC between tau
reader experience (ICCSUVr = 0.62 ± 0.07, ICCR1 = 0.6
Conclusion: Perfusion phase imaging of 18F-PI-2620
injury. Dynamic imaging or a dual time point protoco
PET imaging by indexing both the distribution of tau
Keywords: tau, PET, perfusion, neuronal injury, PI-2

36

n tau-PET perfusion and FDG-PET.
was reached for 18F-PI-2620 acquisition from 0.5 to
r scaling (R > 0.63). Correlation coefficients (summed
nt in all cortical target regions for global mean (RSUVr
ization (RSUVr 0.67 ± 0.10, RR1 = 0.67 ± 0.12). Visual
u-PET perfusion and FDG-PET independent of the
66 ± 0.07).
can serve as a surrogate biomarker for neuronal
ol for tau-PET imaging could replace additional FDG-
and the amount of neuronal injury.
2620

64

Brendel, Matthias

P123: Binding characteristics of 18F-P
predicted tau isoform in suspected 3/4

Matthias Brendel1, Mengmeng Song1, Leonie B
Ken Marek7,8, Carla Palleis9, Lena Kaiser1, Ges
Hammes4, Michael Barbe5, Özgür Onur5, Frank
Schroeter14,15,16, Jost-Julian Rumpf13, Michael R
Bernd Neumaier3,4, Oliver Barret7,8, Jennifer Ma
Stephens17, Sigrun Roeber18, Jochen Herms6,18,
Classen13, Guenter Hoeglinger6,10,19, Peter Barte
Drzezga4,6, John Seibyl7,8, Osama Sabri2

1Department of Nuclear Medicine, University Hospita
2Department of Nuclear Medicine, University of Leipz
3Cognitive Neuroscience, Institute for Neuroscience a
Juelich, Germany
4Department of Nuclear Medicine, University Hospita
5Department of Neurology, University Hospital Colog
6German Center for Neurodegenerative Diseases (DZ
7InviCRO, LLC, Boston, MA, US
8Molecular Neuroimaging, A Division of inviCRO, Ne
9Department of Neurology, University Hospital of Mu
10Department of Neurology, Technical University Mun
11Center for Memory Disorders, University Hospital C
12Department of Psychiatry, University Hospital Colo
13Department of Neurology, University of Leipzig, Lei
14Clinic for Cognitive Neurology, University of Leipzi
15LIFE - Leipzig Research Center for Civilization Dis
16Max- Planck-Institute of Human Cognitive and Brai
17Life Molecular Imaging GmbH, Berlin, Germany
18Center for Neuropathology and Prion Research, Un
Germany
19Department of Neurology, Medizinische Hochschule
20Munich Cluster for Systems Neurology (SyNergy), M
21Department of Molecular Imaging & Therapy, Austi
22The Florey Institute of Neuroscience and Mental He
Australia
23Department of Medicine, Austin Health, The Univer

Aim: Tau proteins consist of different isoforms, chara
microtubule binding sites. Preliminary evidence sugge
18F-PI-2620 is able to visualize the predominantly 3/4
tauopathies Corticobasal syndrome (CBS) and Progre
from the obvious topographical differences - likely wi
them. The aim of this study was to determine whether
between 3/4R- and 4R-tauopathies.

Methods: We evaluated 14 patients with suspected 3/
tauopathy (14 CBS, 15 PSP) at two different centers a
PET scans were acquired 0-60min p.i. and distribution
calculated. Cortical and subcortical clusters exceeding

36

PI-2620 differentiate the clinically
4-repeat and 4-repeat tauopathies

Beyer1, Henryk Barthel2, Thilo van Eimeren3,4,5,6,
sine Respondek6,10, Julia Sauerbeck1, Jochen
k Jessen6,11,12, Dorothee Saur13, Matthias L.
Rullmann2, Andreas Schildan2, Marianne Patt2,
adonia7,8, David S. Russel7,8, Andrew
Kai Bötzel9, Johannes Levin6,9, Joseph
enstein1,20, Victor Villemagne21,22,23, Alexander

al of Munich, LMU Munich, Munich, Germany
zig, Leipzig, Germany
and Medicine (INM-3), Research Centre Juelich,

al Cologne, Cologne, Germany
gne, Cologne, Germany
ZNE), Bonn, Germany

ew Haven, CT, US
unich, LMU Munich, Munich, Germany
nich, Munich, Germany
Cologne, Cologne, Germany
ogne, Cologne, Germany
ipzig, Germany
ig, Leipzig, Germany
seases, University of Leipzig, Leipzig, Germany
in Sciences, Leipzig, Germany

niversity Hospital of Munich, LMU Munich, Munich,

e Hannover, Hannover, Germany
Munich, Germany
in Health, Heidelberg, Australia
ealth, The University of Melbourne, Melbourne,

rsity of Melbourne, Melbourne, Australia

acterized by the number of repeats (R) of their
ests that the novel second-generation tau PET tracer
4R-tauopathy Alzheimer’s disease (AD) and the 4R-
essive supranuclear palsy (PSP) by PET, but - apart
ith different kinetics and magnitude of affinity among
r binding characteristics of 18F-PI-2620 are different

/4R tauopathy and 29 patients with suspected 4R
according to current diagnosis criteria. 18F-PI-2620
n volume ratios (DVR, cerebellar reference) were
g 20 voxels of elevated DVR (≥ 2.5 SD vs. 10 healthy

65

controls) per region were evaluated by non-invasive k
DVR, 30-60 min standard-uptake-value-ratios (SUVr3
min p.i. were compared between 3/4R- and 4R-tauopa
Results: Cortical tau-positive clusters in 4R-tau cases
3/4R-tau cases (p<0.001). Higher mean DVR (1.35±0
(1.67±0.33 vs. 1.31±0.22, p<0.001) and steeper slopes
in cortical clusters of 3/4R-tau cases when compared t
significant differences.
Conclusion: 18F-PI-2620 binding characteristics in tau
suspected 3/4R-tauopathies from 4R-tauopathies. Hig
4R tauopathies when compared to 3/4R-tau cases.
Keywords: tau, PET, isoforms, binding parameter, PI

36

kinetic modelling. R1 (perfusion), k2 (clearance),
30-60) and the linear slope of SUVr between 10 and 60
athies.
s had equal R1 but higher k2 values when compared to
0.19 vs. 1.17±0.60, p<0.01), higher mean SUVr30-60
s (0.85±0.51 vs. 0.43±0.49, p<0.001) were observed
to 4R-tau cases. Subcortical clusters did not show
u-positive cortical regions differentiate clinically
gher tracer clearance indicates less stable binding in
I-2620

66

Brendel, Matthias

P124: 18F-PI2620 tau-PET for assessm
neuropathology in corticobasal syndro

Matthias Brendel1, Carla Palleis2, Catharina Prix
Danek2, Matthias Höllerhage3, Julia Sauerbeck1
Mengmeng Song1, Andrew Stephens4, Henryk B
Alexander Drzezga6,7, Thilo van Eimeren6,7,8,9, V
Robert Perneczky7,14,15,16, Christian Haass7,13,17,

1Department of Nuclear Medicine, University Hospita
2Department of Neurology, University Hospital of Mu
3Department of Neurology, Technical University Mun
4Life Molecular Imaging GmbH, Berlin, Germany
5Department of Nuclear Medicine, University of Leipz
6Department of Nuclear Medicine, University Hospita
7German Center for Neurodegenerative Diseases (DZ
8Cognitive Neuroscience, Institute for Neuroscience a
Juelich, Germany
9Department of Neurology, University Hospital Colog
10Department of Molecular Imaging & Therapy, Austi
11The Florey Institute of Neuroscience and Mental He
Australia
12Department of Medicine, Austin Health, The Univer
13Munich Cluster for Systems Neurology (SyNergy), M
14Department of Psychiatry, University Hospital of M
15Neuroepidemiology and Ageing Research Unit, Scho
Cross Hospital, London, UK
16West London Mental Health NHS Trust, Southhall, U
17Chair of Metabolic Biochemistry, Biomedical Cente
Munich, Munich, Germany
18Department of Neurology, Medizinische Hochschule

Aim: The phenotype of corticobasal syndrome (CBS)
(3/4R)-tau in ~25% of histopathologically evaluated p
PI2620 showed high affinity to 3/4R tau in Alzheimer
pathology. The aim of this study was to investigate 18F

Methods: Twenty-two patients (69±8y) with probable
Armstrong-criteria underwent 18F-PI2620 PET togeth
Distribution volume ratios (DVR, 0-60min) of subcor
cerebellar reference tissue. DVR-data were quantitativ
healthy-controls. Regional 18F-PI2620 binding was co
duration. Amyloid-PET served for assessment of β-am

Results: 23% (5/22) of CBS patients (PSPRS: 24±14)
PI2620 DVR was observed in the whole group of CBS
globus pallidus and subthalamic nucleus while a trend
discernible 18F-PI2620 signal was observed in 14 subj
amyloid negative, 5/5 (100%) of amyloid-positive CB
heterogeneous with involvement of motor and/or parie
PI2620 binding was not associated with disease severi

36

ment of heterogeneous
ome

x2, Mona Gehmeyr2, Kai Bötzel2, Adrian
1, Leonie Beyer1, Alexander Nitschmann1,
Barthel5, Marianne Patt5, Osama Sabri5,
Victor Villemagne10,11,12, Peter Bartenstein1,13,
Johannes Levin2,7, Guenter Hoeglinger3,7,18

al of Munich, LMU Munich, Munich, Germany
unich, LMU Munich, Munich, Germany
nich, Munich, Germany

zig, Leipzig, Germany
al Cologne, Cologne, Germany
ZNE), Bonn, Germany
and Medicine (INM-3), Research Centre Juelich,

gne, Cologne, Germany
in Health, Heidelberg, Australia
ealth, The University of Melbourne, Melbourne,

rsity of Melbourne, Melbourne, Australia
Munich, Germany
Munich, LMU Munich, Munich, Germany
ool of Public Health, Imperial College, Charing

UK
er (BMC), University Hospital of Munich, LMU

e Hannover, Munich, Germany

) is characterized by 4-repeat (4R)-tau in ~50% and by
patients. The next generation tau-PET ligand 18F-
r’s disease (AD) and also revealed affinity to 4R-tau
F-PI2620 in patients with clinical CBS.

e or possible CBS according to MDS-PSP or
her with ten age-matched healthy-controls.
rtical and cortical brain-regions were generated using
vely and visually compared between CBS and
ompared with clinical severity (PSPRS), and disease
myloid status.

) were amyloid-positive. Significantly elevated 18F-
S patients versus healthy controls in the putamen,
d was observed in frontal cortex. Overall, a visually
jects (64%) of the total cohort [9/17 (53%) of the
BS patients]. Cortical binding in CBS was
eto-temporal regions (positive in 41%; 9/22). 18F-
ity but showed a correlation with disease duration in

67

frontal cortex (R=0.48, p=0.039).
Conclusion: Our data indicate a value of 18F-PI2620 f
heterogeneous neuropathology and variable cortical an
target engagement.
Keywords: tau, PET, CBS, heterogeneity, PI-2620

36

for evaluation of CBS, facilitating detection of
nd subcortical deposition sites, which could serve for

68

Bullich Roig, Santiago

P125: Optimal reference region for th
brain using 18F-PI-2620 PET

Santiago Bullich1, Andre Müller1, Núria Roé-Ve
Susan De Santi2, Norman Koglin1, Andrew W. S

1Life Molecular Imaging GmbH, Berlin, Germany
2Life Molecular Imaging Inc, Boston, MA, US
Introduction. Definition of a suitable reference regio
tau PET. Cerebellar cortex is widely used as RR to qu
value ratio (SUVR). However, this RR is not free of p
a small region, close to the edge of the field of view a
regions with tau deposition (e.g. inferior temporal-lob
in the vermis. The objective of this study was to deter
quantification of the tau load using 18F-PI-2620 PET.
Methods. Three healthy controls and three Alzheimer
PI-2620 PET scans. SUVRs were calculated using reg
the standard Automated Anatomic Labeling template
the MRI. Several cerebellar reference regions were us
vermis (RR2), CGM removing vermis and anterior lob
cerebellar superior lobe (RR4). All the reference regio
matter (CWM).
Results. Whole CGM (RR1) provided the lowest test-
increased in smaller RRs (3.70% (RR2), 4.3% (RR3) a
RR2 were lower in comparison to RR3 and RR4 due to
CWM kinetics and activity at pseudo-equilibrium was
including CWM in any RR did not provide significant
Conclusion. CGM excluding vermis and anterior lobe
balance between accuracy and TRTV.
Keywords: tau, PI-2620, reference region

36

he quantification of tau load in the

ellvé1, Aleksandar Jovalekic1, Audrey Perrotin1,
Stephens1

on (RR) is essential for accurate quantification of static
uantify tau PET by means of the standardized uptake
potential drawbacks. Cerebellar gray matter (CGM) is
and with the risk of activity spill-over from nearby
be). Additionally, tau tracers quite often show uptake
rmine the optimal way to delineate cerebellar RR for
r’s disease subjects (58-75 years) underwent two 18F-
gions of interest derived from the intersection between
and the normalized gray matter segmentation from
sed to derive SUVRs: CGM (RR1), CGM removing
be surrounding the vermis (RR3), CGM removing
ons were tested without and with the cerebellar white

-retest variability (TRTV) (3.52% (median)). TRTV
and 4.44% (RR4)). SUVR obtained using RR1 and
o the bias induced by tracer retention in the vermis.
s similar to the activity in the CGM. However,
t improvement in terms of TRTV.
e surrounding the vermis (RR3) provided the optimal

69

Chen, Charles

P126: Quantifying tau PET imaging r
binding

Charles Chen1, Brian Gordon1, Austin McCullo
Dincer1, Shaney Flores1, Sarah Keefe1, Angela P
Yi Su2, John Morris1, Tammie Benzinger1

1Washington University in St. Louis, St. Louis, MO, U
2Banner Health, Phoenix, AZ, US
Introduction: Tau PET imaging is a powerful tool fo
tangles across the brain, but off-target binding raises q
uptake indicates a positive scan for tau pathology in A
on the distribution of elevated tracer uptake across the
based on the pattern of neurofibrillary lesions at autop
tracer uptake in temporal lobe regions. While both typ
tau PET biomarkers for Alzheimer’s disease, they rely
individuals grouped by amyloid PET imaging and cog
instrumental in avoiding complications from off-targe
complications when translating tau PET imaging into
Methods: To avoid these complications, we propose a
and a threshold for tracer uptake solely from the tau P
from normal aging to Alzheimer’s disease (n = 388).
Results: We evaluate model performance by testing h
rate = 4.58%, n = 131), how well the model separates
Alzheimer’s disease, Cliff’s d = 0.49, preclinical versu
n = 330), and how well the model compares to previou
Discussion: Our results suggest a tau PET biomarker
reliably in the presence of off-target binding.

37

reliably in the presence of off-target

ough1, Aiad Zaza1, Christopher Mejias1, Aylin
Paulick1, Kelley Jackson1, Deborah Koudelis1,

US

or studying the in vivo pattern of neurofibrillary
questions about where and what degree of tracer
Alzheimer’s disease. One common approach focuses
e brain by using a modified Braak staging, which is
psy. Other approaches focus more on the degree of
pes of approaches have been pivotal in establishing
y on heuristic thresholds for tracer uptake in
gnitive testing results; while these groupings may be
et binding, their a priori assumptions could raise new
future clinical use.
a probabilistic model that learns regions of interest
PET imaging data of a cohort spanning the spectrum

how well the model agrees with visual reads (error
clinical groups (controls versus preclinical
us symptomatic Alzheimer’s disease, Cliff’s d = 0.63,
us approaches for quantifying tau positivity.
learned solely from tau PET imaging data performs

70

Keywords: Tau PET, Alzheimer’s disease, probablisti
37

ic modeling, off-target binding, visual reads
71

Chiotis, Konstantinos

P127: [18F]THK5317 imaging as a pre
decline in prodromal and dementia-sta

Konstantinos Chiotis1,2, Irina Savitcheva3, Kons
Anders Wall6,7, Gunnar Antoni7, Agneta Nordbe

1Nordberg Translational Molecular Imaging Lab, Div
Research, Department of Neurobiology, Care Science
Sweden
2Theme Neurology, Karolinska University Hospital, S
3Medical Radiation Physics and Nuclear Medicine, K
4Westman neuroimaging group, Division of Clinical G
Department of Neurobiology, Care Sciences and Soci
5Toulouse NeuroImaging Center, University of Toulou
6Department of Surgical Sciences, Uppsala University
7Department of Medicinal Chemistry, Uppsala Univer
8Theme Aging, Karolinska University Hospital, Stockh

Background: Tau-PET imaging has proved promisin
impairment in Alzheimer’s disease (AD) in cross-sect
baseline [18F]THK5317 binding to predict prospectiv
compared the tracer’s accuracy with that of [18F]FDG
markers.

Methods: Twenty patients with clinical biomarker-ba
impairment and a positive amyloid-beta [11CPIB] PE
underwent at baseline neuropsychological assessment
glucose metabolism, CSF-sampling, and structural MR
range=24.88:43.70) months, the patients underwent fo
cognitive decline (cognitively stable vs cognitive dete

Results: Eleven patients deteriorated further cognitive
up. Baseline [18F]THK5317 binding load showed exc
under the curve (AUC) 0.96-1.00) and related strongly
the other baseline biomarkers showed fair/poor accura
did not associate with the pace of decline. Baseline [1
stronger with MMSE at follow-up than at baseline.

Conclusions: These findings support the temporal off
and highlight that tau-PET performs better than other
decline. Our results question the deterministic role of
classifying patients in the AD-continuum of cognitive
Tau-PET could prove pertinent for accurate diagnostic

37

edictive tool of prospective cognitive
age Alzheimer’s disease

stantinos Poulakis4, Laure Saint-Aubert1,5,
erg1,8

vision of Clinical Geriatrics, Center for Alzheimer
es and Society, Karolinska Institutet, Stockholm,

Stockholm, Sweden
Karolinska University Hospital, Stockholm, Sweden
Geriatrics, Center for Alzheimer Research,
iety, Karolinska Institutet, Stockholm, Sweden

use, Inserm, UPS, Toulouse, France
y, Uppsala, Sweden
rsity, Uppsala, Sweden
holm, Sweden

ng in following neurodegeneration and cognitive
tional designs. This study assessed the accuracy of
ve cognitive decline in patients with AD and
G PET, CSF, neuropsychological and atrophy

ased diagnosis in the AD-spectrum (cognitive
ET scan; prodromal and dementia-stage AD)
t, PET with [18F]THK5317 for tau, [18F]FDG for
RI. After a median of 35.62 (interquartile
ollow-up assessment for evaluating clinically evident
erioration groups), including MMSE rating.

ely, while nine remained cognitively stable at follow-
cellent accuracy at predicting cognitive decline (area
y with the pace of decline. Conversely, the levels of
acy to predict cognitive decline (AUC 0.58-0.77) and
18F]THK5317 binding and CSF tau associated

fset between tau deposition and cognitive impairment,
biomarkers in predicting prospective cognitive
a single amyloid-beta biomarker positivity for
e decline at clinically-relevant follow-up intervals.
cs and as a screening tool in clinical trials.

72

(A) Voxel-wise comparisons,
(B) sensitivity and
(C) specificity in discriminating between patients exh
deteriorated cognitively over time, based on baseline [
binding/uptake.
Keywords: Tau PET, amyloid PET, CSF, follow-up, c

37

hibiting cognitive stability and patients that
[18F]THK5317, [18F]FDG and [11C]PIB
cognition

73

Cohen, Ann

P128: Comparison of quantitative cuto

Davneet Minhas1, Charles Laymon1, Brian Lopr
Aizenstein4, Oscar Lopez2, Chester Mathis1, Wi

1University of Pittsburgh, School of Medicine, Depart
2University of Pittsburgh, School of Medicine, Depart
3University of Pittsburgh, School of Medicine, Depart
4University of Pittsburgh, School of Medicine, Depart

Background: [18F]AV-1451 has allowed for assessme
However, a gap exists in classification of [18F]AV-145
here is to evaluate quantitative methods of [18F]AV-14
consensus visual reads of [18F]AV-1451 images.

Methods: [18F]AV-1451 scans were collected in 211
clinical diagnoses (5AD, 37MCI, 169NC). Braak stag
using FreeSurfer, and regional SUVR measures and im
We applied two quantitative methods for determining
the iterative outlier (IO) approach and k-means (KM)
visual reads of [18F]AV-1451 SUVR images for 76 su
agreement between visual reads and the two cutoff me

Results: Overall, IO demonstrated higher, more conse
methods, there was relatively good agreement betwee
agreement decreased for KM with increasing Braak st
Braak stages (Table 1).

Discussion: The KM and IO approaches generally agr
KM, particularly in higher Braak regions. Additionall
unreasonably low, an explanation for this phenomeno
used here, KM may have failed to find suitable midpo
agreement between visual reads and quantitative appr
composite regions and smaller positive hotspots withi
will explore the performance of these methods across
performance.

Keywords: AV-1451, cutoffs

37

off methods for [18F]AV-1451

resti1, Beth Snitz2, Dana Tudorascu3, Howard
iliam Klunk4, Ann Cohen4

tment of Radiology, Pittsburgh, PA, US
tment of Neurology, Pittsburgh, PA, US
tment of Medicine, Pittsburgh, PA, US
tment of Psychiatry, Pittsburgh, PA, US
ent of brain tau pathology across the AD spectrum.
51 into tau-positive and -negative groups. The goal
451 classification and compare these methods to

1 elderly subjects (78M, 133F, 75.7±6.3) across
ge (I,III-VI) regions-of-interest (ROIs) were defined
mages were created using cerebellar gray as reference.
g [18F]AV-1451 Braak ROI cutoffs to the entire cohort,

cluster approach. We then performed consensus
ubjects (34M, 42F, 76.1±71), and assessed overall
ethods.
ervative cutoffs across Braak stages. For both
en visual reads and the cutoffs for Braak 1. This
tage while IO showed better agreement in higher

reed with visual reads. IO cutoffs were higher than
ly, KM cutoffs in higher Braak regions were
on may be because of the small number of AD patients
oints in higher Braak regions. Additionally, lower
roaches may be because Braak ROIs are large
in higher Braak ROIs may be diluted. Ongoing studies
all FreeSurfer regions, to further evaluate their

74

Fernandez Arias, Jaime

P129: Pilot study on the relationship b
early and late stages of AD

Jaime Fernandez Arias1,4, Tharick Pascoal1,4, An
Kang1,3,4, Melissa Savard1,4, Julie Ottoy2, Sulant
Tissot1,4, Emilie Thomas1, Jenna Stevenson1,4, N
Massarweh3, Jean-Paul Soucy3, Serge Gauthier1

1Translational Neuroimaging Laboratory, The McGill
Montreal, QC, Canada
2Molecular Imaging Center Antwerp, University of An
3Montreal Neurological Institute, Montreal, QC, Cana
4Douglas Mental Health Research Institute, Verdun, Q

Introduction: multiple studies have linked the develo
neurodegeneration in AD. A few studies have previou
morphometry (VBM) and PET using first-generation
our knowledge, no study has reported this association
tracer with higher sensitivity to tau tangles as compare

Objective: we aimed at exploring the relationship bet
VBM in the Translational Biomarkers in Aging and D

Methods: Structural MRI and tau PET (18F-MK-624
impairment (MCI) and 19 young control participants.
and white (WM) maps, non-linearly registered to the A
8mm FWHM gaussian kernel. Voxel-wise linear mod
association between VBM and tau PET adjusting for a
whole sample and within diagnostic groups.

Results: we found a negative association between VB
MCI or AD in the medial temporal lobe (MTL), the te
parietal cortex and the left prefrontal cortex (PFC). W

Conclusions: the pilot study further confirms previou
and neurodegeneration in brain areas susceptible to A
MK-6240 as a promising biomarker for tau pathology

37

between 18F-MK-6240 and VBM in

ndrea Benedet1,4, Joseph Thierrault1,4, Min Su
tha Mathotaarachchi1, Firoza Lussier1,4, Cécile
Nesrine Rahmouni1,4, Tina Wang1,4, Gassan
1,3,4, Pedro Rosa-Neto1,3,4

l University Research Centre for Studies in Aging,
ntwerp, Antwerp, Belgium
ada
QC, Canada
opment of tau tangles to neuronal loss or
usly examined the association between Voxel based
tracers, mostly 18F-AV-1451. However, to the best of
n using 18F-MK-6240, which is a second-generation
ed to 18F-AV-1451.
tween tau PET, indexed here by 18F-MK-6240, and
Dementia (TRIAD) cohort.
40) were acquired for 34 AD, 47 mild cognitive
MRI were segmented into probabilistic grey (GM)
ADNI template using Dartel and smoothed with an
dels were applied, using VoxelStats, to test for the
age and gender. Analyses were performed in the

BM and tau uptake in participants with a diagnose of
emporoparietal cortices, the occipital cortex, the
We did not find any association in young controls.
us findings on the association between tau pathology
AD pathology. In addition, our study supports 18F-
y.

75

Keywords: tau, PET, VBM, 18F-MK-6240

37