Thesis

Optimizing the use of dynamic tau PET in Alzheimer’s Disease methodological and clinical applications Denise Visser

Optimizing the use of dynamic tau PET in Alzheimer’s Disease: methodological and clinical applications Denise Visser

The research described in this thesis was carried out at the department of Radiology & Nuclear Medicine and Alzheimer Center Amsterdam, Amsterdam UMC, Amsterdam, The Netherlands. Research of the Alzheimer Center is part of the Neurodegeneration program of Amsterdam Neuroscience. The Alzheimer Center is supported by Stichting Alzheimer Nederland and Stichting Steun Alzheimercentrum Amsterdam. Tau-PET research in this thesis was made possible by Avid Radiopharmaceuticals, ZonMW Memorabel and Stichting Alzheimer Nederland. Printing of this thesis was supported by Stichting Alzheimer Nederland, Stichting Alzheimer en Neuropsychiatrie and Vrije Universiteit Amsterdam. Cover concept: Leo & Denise Visser Cover design: Designed by Loniek Layout: Bart Roelofs, persoonlijkproefschrift.nl Printed by Ipskamp Printing | proefschriften.net ISBN: 978-94-6473-321-1 © 2023 Denise Visser All rights reserved. Save expectations stated by law, no part of this thesis may be reproduced stored in a retrieval system of any nature, or transmitted, in whole or in part, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, included in a complete or partial transcription, without the written permission of the author, or- when appropriate – of the publisher’s of the presented publications.

Voor oma

TABLE OF CONTENTS Chapter 1 Introduction 9 Part I: Methodological considerations 21 Chapter 2 Effect of Shortening the Scan Duration on Quantitative Accuracy of [18F]Flortaucipir Studies 23 Molecular Imaging and Biology (2021) Chapter 3 Longitudinal tau PET using 18F-flortaucipir: the Effect of relative Cerebral Blood Flow on (semi)quantitative parameters 49 The Journal of Nuclear Medicine (2023) Part IIa: Clinical applications – cognitive correlates 77 Chapter 4 Tau pathology and relative cerebral blood flow are independently associated with cognition in Alzheimer’s disease 79 European Journal of Nuclear Medicine and Molecular Imaging (2020) Chapter 5 Differential associations between neocortical tau pathology and blood flow with cognitive deficits in earlyonset vs late-onset Alzheimer’s disease 109 European Journal of Nuclear Medicine and Molecular Imaging (2022) Chapter 6 Changes in tau PET and relative cerebral blood flow as predictors of longitudinal cognitive trajectories on the Alzheimer’s continuum 141 Submitted Part IIb: Clinical applications – biomarkers 177 Chapter 7 Tau pathology as determinant of changes in atrophy and cerebral blood flow: a multi‑modal longitudinal imaging study 179 European Journal of Nuclear Medicine and Molecular Imaging (2023) Chapter 8 Longitudinal change in ATN biomarkers in cognitively normal individuals 205 Alzheimer’s Research & Therapy (2022) Chapter 9 Summary & Discussion 227 Appendices Nederlandse samenvatting 243 List of publications 257 List of author affiliations 261 List of theses of the Alzheimer Center Amsterdam 264 PhD portfolio 269 Dankwoord 270 About the author 275

1

Introduction

10 Chapter 1 INTRODUCTION Dementia and Alzheimer’s disease Dementia represents a worldwide public health problem, with more than 55 million people living with dementia (World Health Organization, March 2023; Dementia (who.int)). This number is expected to increase up to 131 million people living with dementia by 2050 [1]. The most common cause of dementia (~70% of all cases) is Alzheimer’s disease (AD). The development of symptoms in AD takes decades and can be characterized by several stages describing progression of impairment [Figure 1]. The preclinical AD stage refers to presence of AD pathology in the absence of objectively measurable impairment. This stage also includes subjective cognitive decline (SCD), which refers to the subjective experience of cognitive decline, without objective impairment on cognitive assessment. Compared to cognitively normal individuals who do not experience complaints, individuals with SCD have an increased risk of subsequent objective cognitive decline, especially in a memory clinic setting [2-4]. As such, SCD can be a possible first expression of AD, albeit SCD can have many causes and most individuals with SCD do not harbour AD pathology. The next stage, in which cognitive decline is formally objectified (by e.g. neuropsychological assessment), is called mild cognitive impairment (MCI) due to AD [5]. In individuals with MCI due to AD, there is impairment, but cognitive decline does not yet interfere substantially with everyday activities. The stage in which the presence of progressive loss of cognition in multiple cognitive domains co-occurs with interference in daily functioning (eg, social, occupational, self-care) is called AD dementia. Depending on the degree of interference with daily life activities, the AD dementia stage can be subdivided into mild, moderate, and severe AD dementia [Figure 1]. While AD dementia most often affects individuals of older age, a minority of individuals develop symptoms at a younger age (usually defined as below the age of 65), referred to as early onset AD (EOAD). From a clinical perspective, AD dementia is typically characterized by amnestic symptoms (memory impairment), with often co-occurrence of non-amnestic symptoms (e.g. language or executive functioning impairment or behavioral changes), especially in later stages of the disease. When non-amnestic symptoms characterize the clinical phenotype early in the disease, it can be referred to as an atypical variant AD [6]. Atypical variants of AD affect young people (EOAD) disproportionally often. The presence of non-amnestic symptoms leads to diagnostic challenges, as they can closely resemble symptoms related to other causes of dementia. For an accurate diagnosis, prognosis and health care management plan it therefore is essential to gain insight into the underlying pathophysiology of the disease.

11 Introduction Figure 1. Clinical stages of AD (Histo)pathology of Alzheimer’s disease and the role of PET In general, AD pathological changes start with extracellular deposition of fibrillar amyloid-β (plaques), followed by intracellular accumulation of hyperphosphorylated tau (in neurofibrillary tangles; tau pathology), synaptic and neuronal dysfunction (among which altered cerebral blood flow (CBF)), neurodegeneration, cognitive dysfunction and, finally, dementia [7, 8]. The first pathological changes in AD develop already an estimated 15-20 year prior to symptoms onset. Formation of amyloid-β containing plaques is considered (one of the) earliest pathological events in AD, most often firstly present in the (orbito)frontal cortex and precuneus [Figure 2]. Already in early disease stages amyloid pathology spreads throughout the rest of the brain, with very limited regional specificity and little between-person variability. Other than amyloid pathology, accumulation of hyperphosporylated tau proteins does show large variability in distribution. Notwithstanding, the development generally seems to occur according to a specific topographical pattern [9, 10]. This pattern is referred to by ‘Braak stages’, describing accumulation starting in transentorhinal regions (Braak stage I), followed by hippocampal (Braak stage II) and limbic regions (Braak stage III/IV), and eventually neocortical association cortices (Braak stage V/VI) [Figure 2] [10]. Whereas tau pathology in the medial temporal lobe can be observed in cognitively unimpaired individuals, cognitively impaired individuals (almost) always show tau pathology in the neocortex [10]. 1

12 Chapter 1 Figure 2. Neurofibrillary tangle (tau pathology) and plaque (amyloid pathology) development in AD (adapted from Van der Kant et al. Nature Reviews (2020) [9]). Whereas an AD diagnosis used to be based on clinical symptomatology, nowadays there is a shift towards a neurobiological diagnosis of the disease. The AT(N) classification provides a framework to diagnose AD based on biomarkers providing an indication of these pathologic changes [11]. In this framework, individuals are not only classified by the presence or absence of amyloid (plaques, ‘A’) and hyperphosphorylated tau (neurofibrillary tangles, ‘T’), but also by neurodegeneration or neuronal injury (‘N’). Biomarker status can be determined using positron emission tomography (PET). PET is a molecular imaging technique that allows the visualization of many physiological processes in vivo. Visualizing AD pathology in vivo became possible in 2004 with the introduction of the amyloid-specific PET tracer [11C]Pittsburgh Compound-B, also known as [11C]PIB [12-14]. Since then, amyloid PET has been researched extensively, multiple amyloid PET tracers have been developed and nowadays several of these amyloid PET tracers are approved for clinical use. Although amyloid PET can be used to determine ‘A’ status in the AT(N) classification system, experience with amyloid PET has shown that, apart from its limited regional specificity and between-person variability, amyloid PET does not (or maximally very minimally) associate with cognitive performance in AD. In contrast, tau pathology does show regional specificity and large between-person variability, but more importantly, tau pathology associates with cognitive performance in AD. This makes imaging of tau pathology in vivo of great interest. PET imaging of tau pathology became possible only several years ago, and since the first PET tracers for (AD-specific) tau pathology became available (first-generation tau PET tracers), nowadays multiple tau PET tracers (both first and second generation) are available for research purposes [15]. These tracers bind specifically to the AD-specific tau

13 Introduction pathology (i.e., a combination of 3R/4R isoforms). The most widely used tracer for AD-specific tau pathology is [18F]flortaucipir. This tracer captures the paired helical filament of tau in neurofibrillary tangles. In Europe [18F]flortaucipir is currently only approved for research purposes, but In America the tracer has been approved for clinical use by the American Food & Drug Agency (FDA) in 2020 [16]. Tau PET can be used as a measure for tau pathology in the AT(N) classification, where ‘T’ status can be based on either quantitative thresholds or visual read. There are only few studies evaluating the temporal ordering of AT(N) biomarker abnormality in a longitudinal manner, using [18]flortaucipir for determination of ‘T’ status. Insight into the temporal dynamics of the biomarkers used in the diagnostic framework is important, especially for the earliest phases of disease in which pathological changes might direct future prevention or treatment strategies. Methodological aspects of Tau PET PET images can be acquired using dynamic or static protocols. Static scanning protocols consist of tracer injection, followed by a waiting period, after which a short scan is acquired. Static protocols have the advantage of clinical applicability and relative computational simplicity, but harbor the disadvantage that the parameters obtained are semi-quantitative. Dynamic PET scan protocols consist of tracer injection, while simultaneously starting a long scanning acquisition. Dynamic protocols allow for more accurate (fully) quantitative measures of specific binding of PET tracers [17, 18]. Moreover, dynamic scan protocols additionally enable computation of parametric images of tracer delivery, which can be interpreted as a proxy of relative tracer flow or relative cerebral blood flow (rCBF) (i.e. R1) [19]. R1 represents the ratio between the rate constant for ligand transfer from blood to tissue (K1) in the tissue of interest and the reference region, which is strongly correlated with metabolic activity ([18F]FDG PET) and the ‘gold standard’ for measuring flow ([15O]H 2O PET) [20-22]. A dynamic [18F]flortaucipir PET scan may thus not only provide accurate information on (regional) quantification of tau pathology, but also yields information on rCBF. Semi-quantitative measures from static scans are usually sufficient for clinical application, but accurate quantification of tracer uptake is of major importance for accurately detecting early-stage pathology, clinical trials and longitudinal studies. Longitudinal changes using static measures may be biased by blood flow changes, whereas quantitative measures are not. However, the PET tracer [18F]flortaucipir requires a relatively long dynamic acquisition period (i.e. up to 100 minutes) because of the slow tracer kinetics. This can be challenging, especially when working with a vulnerable population (like individuals with AD), and even more so when applied in a longitudinal or clinical trial setting. Ideally, one would thus prefer a relatively short dynamic scanning protocol, yielding quantitative accurately information about 1

14 Chapter 1 both tau pathology and cerebral blood flow, while being more easily applicable in AD populations and longitudinal settings. However, whether it is feasible to shorten the dynamic tau PET scanning protocol while maintaining quantitative accuracy is yet to be determined. Furthermore, measuring R1 in longitudinal studies is especially important because blood flow changes can occur over time in AD because of disease progression or drug intervention. However, for [18F]flortaucipir the sensitivity of static measures for changes in blood flow has not been investigated and it thus remains unclear whether longitudinal [18F]flortaucipir PET studies require dynamic tracer acquisition in the context of blood flow induced bias. Biological and clinical correlates of Tau PET In contrast to amyloid-β deposition, histopathological studies have shown tau pathology correlates well with cognitive impairment and its topography is specific to the different AD clinical phenotypes (e.g. tau pathology predominantly in the occipital visual association cortex in the visual variant of AD or frontal tau pathology in the behavioral variant of AD) [23]. Some first PET studies demonstrated that high levels of regional tau PET levels [10–12], as well as low levels of rCBF (as measured with [18F]FDG PET or MRI), correlate with cognitive impairment in various domains. However, more studies in different cohorts are required to improve understanding about the relationship between tau PET and cognition in AD. Furthermore, rCBF has not been investigated yet using [18F]flortaucipir R 1 prior to this thesis. Investigating tau pathology and rCBF simultaneously by using dynamic [18F]flortaucipir PET might yield valuable information, since both pathophysiological mechanisms may contribute to cognitive impairment in AD and little is known about their interrelationships. Since clinical (cognitive) phenotypes depend on age-at-onset, clinical phenotypes may also be associated with differences in the regional distribution of tau pathology. Age-at-onset is thus closely related to both cognitive symptoms and distinct patterns of tau pathology in AD. This might imply that there could be differences in the association between tau pathology and cognitive performance depending on age-at-onset. Understanding these potential differences between EOAD and LOAD might have important implications for clinical trials, among others, since effects of potential tau- or blood flow targeting therapeutic interventions might exert different effects on cognition in EOAD compared to LOAD patients. Whether the association between tau pathology and cognitive performance is affected by age-at-onset is yet to be elucidated. Similarly, it is unknown whether age-at-onset affects the association between rCBF and cognitive performance. Another important question in the context of clinical use, but also for clinical trials, would be whether tau PET could serve as a predictive tool for future rates of cognitive decline. This would not only be useful in research context, directing

15 Introduction participant selection for clinical trials, but may also provide patients with a better perspective on what to expect regarding disease progression and related clinical impairment (which is currently lacking). When it comes to using tau PET as a predictive tool, this technique might also forecast neuronal injury. In line with histopathological and in vitro studies, neuroimaging studies have demonstrated strong correlations between baseline tau PET with cross-sectional atrophy in AD patients. In addition, longitudinal studies with relatively short follow-up time (i.e., 12-15 months) have shown that tau load also predicts future atrophy rates. Another aspect of neuronal injury in AD is the progressive reduction of cerebral blood flow (CBF). Previous studies indicate that baseline tau pathology is related to neuronal injury as reflected by decreased CBF as well as atrophy in AD. However, it is less well known whether (rate of) change in tau pathology also relates to (rate of) change in CBF. To better understand how tau PET and neurodegeneration are related or whether tau PET can serve as a predictive tool for future neuronal injury, it is important to study their dynamic associations over time and investigate how baseline tau load and change in tau load associate with longitudinal cortical thinning and rCBF. Aims and outline of this thesis This thesis had two general aims. First, we aimed to gain better understanding of methodological aspects of tau PET in the context of AD, in order to optimize its protocols and applications in clinical and research settings. The second aim of this thesis was to provide insight into the biological and clinical correlates of tau PET in AD, in order to optimize and direct its use in clinical (and prognostics) and research settings. The outline of this thesis starts with addressing the first aim, examining methodological aspects of tau PET in the context of AD. We first aimed to investigate whether a shorter overall scan duration for [18F]flortaucipir PET dualtime-window scans is feasible, while retaining quantitative accuracy (chapter 2). In chapter 3, we compared semi-quantitative (SUVr) and quantitative (DVR/BPND) parameters for [18F]flortaucipir PET in a longitudinal setting. In the second part of this thesis we examined aspects of clinical applications of tau PET. We investigated the (regional) association between tau pathology and rCBF, and their (independent) associations with cognitive functioning in patients with AD (chapter 4). Following this, we investigate differences between early-onset AD and late-onset AD in (1) tau pathology and rCBF and (2) the associations between tau pathology and rCBF with cognitive performance (chapter 5). In chapter 6, we extended into longitudinal applications and assessed the association between i) baseline tau pathology and rCBF, and ii) change in tau pathology and rCBF with longitudinal cognitive change in individuals along the AD spectrum. Expanding on 1

16 Chapter 1 biomarker associations, in chapter 7 we investigated the associations between changes in tau PET with imaging biomarkers of neuronal injury (i.e., atrophy and CBF) in a cohort comprising i) amyloid negative (Aβ-) cognitively normal (CN) individuals, and ii) amyloid positive (Aβ+) CN and cognitively impaired (CI) (AD-phenotype) individuals. Lastly, we identified determinants of change in AT(N) profiles over time in cognitively normal individuals by using tau PET in chapter 8. This thesis will be concluded by a summary and discussion of our findings and their implications.

17 Introduction REFERENCES 1. Prince MJ, Wimo A, Guerchet MM, Ali GC, Wu Y-T, Prina M. World Alzheimer Report 2015The Global Impact of Dementia: An analysis of prevalence, incidence, cost and trends. 2015. 2. Slot RE, Sikkes SA, Berkhof J, Brodaty H, Buckley R, Cavedo E, et al. Subjective cognitive decline and rates of incident Alzheimer’s disease and non–Alzheimer’s disease dementia. Alzheimer’s & Dementia. 2019;15:465-76. 3. Snitz BE, Wang T, Cloonan YK, Jacobsen E, Chang C-CH, Hughes TF, et al. Risk of progression from subjective cognitive decline to mild cognitive impairment: The role of study setting. Alzheimer’s & Dementia. 2018;14:734-42. 4. Jessen F, Amariglio RE, Van Boxtel M, Breteler M, Ceccaldi M, Chételat G, et al. A conceptual framework for research on subjective cognitive decline in preclinical Alzheimer’s disease. Alzheimer’s & dementia. 2014;10:844-52. 5. Albert MS, DeKosky ST, Dickson D, Dubois B, Feldman HH, Fox NC, et al. The diagnosis of mild cognitive impairment due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Focus. 2013;11:96-106. 6. Graff-Radford J, Yong KX, Apostolova LG, Bouwman FH, Carrillo M, Dickerson BC, et al. New insights into atypical Alzheimer’s disease in the era of biomarkers. The Lancet Neurology. 2021;20:222-34. 7. Hardy JA, Higgins GA. Alzheimer’s disease: the amyloid cascade hypothesis. Science. 1992;256:184-5. 8. Selkoe DJ, Hardy J. The amyloid hypothesis of Alzheimer’s disease at 25 years. EMBO molecular medicine. 2016;8:595-608. 9. van der Kant R, Goldstein LS, Ossenkoppele R. Amyloid-β-independent regulators of tau pathology in Alzheimer disease. Nature Reviews Neuroscience. 2020;21:21-35. 10. Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta neuropathologica. 1991;82:239-59. 11. Jack Jr CR, Bennett DA, Blennow K, Carrillo MC, Dunn B, Haeberlein SB, et al. NIA-AA research framework: toward a biological definition of Alzheimer’s disease. Alzheimer’s & Dementia. 2018;14:535-62. 12. Mathis CA, Bacskai BJ, Kajdasz ST, McLellan ME, Frosch MP, Hyman BT, et al. A lipophilic thioflavin-T derivative for positron emission tomography (PET) imaging of amyloid in brain. Bioorganic & medicinal chemistry letters. 2002;12:295-8. 13. Klunk WE, Wang Y, Huang G-f, Debnath ML, Holt DP, Shao L, et al. The binding of 2-(4′-methylaminophenyl) benzothiazole to postmortem brain homogenates is dominated by the amyloid component. Journal of Neuroscience. 2003;23:2086-92. 14. Klunk WE, Engler H, Nordberg A, Wang Y, Blomqvist G, Holt DP, et al. Imaging brain amyloid in Alzheimer’s disease with Pittsburgh Compound‐B. Annals of Neurology: Official Journal of the American Neurological Association and the Child Neurology Society. 2004;55:30619. 1

18 Chapter 1 15. Leuzy A, Chiotis K, Lemoine L, Gillberg P-G, Almkvist O, Rodriguez-Vieitez E, et al. Tau PET imaging in neurodegenerative tauopathies—still a challenge. Molecular psychiatry. 2019;24:1112-34. 16. Barthel H. First tau PET tracer approved: toward accurate in vivo diagnosis of Alzheimer disease. Soc Nuclear Med; 2020. p. 1409-10. 17. Lammertsma AA. Forward to the past: the case for quantitative PET imaging. Journal of Nuclear Medicine. 2017;58:1019-24. 18. Ossenkoppele R, Prins ND, Van Berckel BN. Amyloid imaging in clinical trials. Alzheimer’s research & therapy. 2013;5:1-3. 19. Joseph-Mathurin N, Su Y, Blazey TM, Jasielec M, Vlassenko A, Friedrichsen K, et al. Utility of perfusion PET measures to assess neuronal injury in Alzheimer’s disease. Alzheimer’s & Dementia: Diagnosis, Assessment & Disease Monitoring. 2018;10:669-77. 20. Ottoy J, Verhaeghe J, Niemantsverdriet E, De Roeck E, Ceyssens S, Van Broeckhoven C, et al. 18F-FDG PET, the early phases and the delivery rate of 18F-AV45 PET as proxies of cerebral blood flow in Alzheimer’s disease: Validation against 15O-H2O PET. Alzheimer’s & Dementia. 2019;15:1172-82. 21. Peretti DE, Vállez García D, Reesink FE, Doorduin J, de Jong BM, De Deyn PP, et al. Diagnostic performance of regional cerebral blood flow images derived from dynamic PIB scans in Alzheimer’s disease. EJNMMI research. 2019;9:1-9. 22. Peretti DE, Vállez García D, Reesink FE, van der Goot T, De Deyn PP, de Jong BM, et al. Relative cerebral flow from dynamic PIB scans as an alternative for FDG scans in Alzheimer’s disease PET studies. PloS one. 2019;14:e0211000. 23. Ossenkoppele R, Schonhaut DR, Schöll M, Lockhart SN, Ayakta N, Baker SL, et al. Tau PET patterns mirror clinical and neuroanatomical variability in Alzheimer’s disease. Brain. 2016;139:1551-67.

Part I

Methodological considerations

2

Efffect of Shortening the Scan Duration on Quantitative Accuracy of [18F]Flortaucipir Studies Hayel Tuncel, Denise Visser, Maqsood Yaqub, Tessa Timmers, Emma E. Wolters, Rik Ossenkoppele, Wiesje M. van der Flier, Bart N.M. van Berckel, Ronald Boellaard, Sandeep S.V. Golla Mol Imaging Biol 23, 604–613 (2021)

24 Chapter 2 ABSTRACT Purpose Dynamic positron emission tomography (PET) protocols allow for accurate quantification of [18F]flortaucipir-specific binding. However, dynamic acquisitions can be challenging given the long required scan duration of 130 min. The current study assessed the effect of shorter scan protocols for [18F]flortaucipir on its quantitative accuracy. Procedures Two study cohorts with Alzheimer’s disease (AD) patients and healthy controls (HC) were included. All subjects underwent a 130-min dynamic [18F]flortaucipir PET scan consisting of two parts (0–60/80–130 min) post-injection. Arterial sampling was acquired during scanning of the first cohort only. For the second cohort, a second PET scan was acquired within 1–4 weeks of the first PET scan to assess test-retest repeatability (TRT). Three alternative time intervals were explored for the second part of the scan: 80–120, 80–110 and 80–100 min. Furthermore, the first part of the scan was also varied: 0–50, 0–40 and 0–30 min time intervals were assessed. The gap in the reference TACs was interpolated using four different interpolation methods: population-based input function 2T4k_VB (POP-IP_2T4k_VB), cubic, linear and exponential. Regional binding potential (BPND) and relative tracer delivery (R1) values estimated using simplified reference tissue model (SRTM) and/or receptor parametric mapping (RPM). The different scan protocols were compared to the respective values estimated using the original scan acquisition. In addition, TRT of the RPM BPND and R1 values estimated using the optimal shortest scan duration was also assessed. Results RPM BPND and R1 obtained using 0–30/80–100 min scan and POP-IP_2T4k_VB reference region interpolation had an excellent correlation with the respective parametric values estimated using the original scan duration (r2 > 0.95). The TRT of RPM BP ND and R1 using the shortest scan duration was − 1 ± 5 % and − 1 ± 6 % respectively. Conclusions This study demonstrated that [18F]flortaucipir PET scan can be acquired with sufficient quantitative accuracy using only 50 min of dual-time-window scanning time.

25 Shortening the scan duration INTRODUCTION Dynamic positron emission tomography (PET) scan protocols allow for accurate quantitative measures [1, 2] of specific binding of PET tracers. Moreover, dynamic scan protocols yield additional information about functional measures such as perfusion [3]. Semi-quantitative measures from static scans are usually sufficient for clinical application, but accurate quantification of tracer uptake is of major importance in the context of early-stage pathology, clinical trials [1] and longitudinal studies. Some PET tracers like the tau tracer [18F]flortaucipir require a long acquisition period because of the slow tracer kinetics. This can be challenging, especially when working with a vulnerable population (like patients with Alzheimer’s disease (AD)). In vivo quantification of tau pathology is important because intracellular accumulation of hyperphosphorylated tau proteins into neurofibrillary tangles (NFTs) is one of the pathological hallmarks of AD [4]. Indeed, histopathological studies have shown that the amount of NFTs correlate well with the severity of their cognitive symptoms during life [5, 6]. [18F]Flortaucipir is worldwide the most widely used PET tracer for detecting and quantifying these NFTs. For the analysis of [18F]flortaucipir scans, most studies prefer semi-quantitative measures due to their practical applicability and computational simplicity [7–9]. However, studies involving dynamic imaging provided more accurate and precise pharmacokinetic parameters and provide estimates for relative tracer delivery (R1) or relative cerebral blood flow (rCBF) [2, 10–15], which is important for monitoring flow changes. For instance, a study by van Berckel et al. [16] observed that longitudinal changes in [11C]PIB standardized uptake value ratio (SUVr) do not reflect changes in specific [11C]PIB binding but rather are secondary to changes in blood flow during the natural course of AD. Our group has performed dynamic acquisition of [18F]flortaucipir scans, using a 130-min dual-time-window dynamic scan protocol including a 20-min break (after the first 60 min of acquisition) [17–21]. Several aspects are of importance to obtain a reliable protocol with reduced overall scanning time. Firstly, the scan must include the wash-in of the tracer and tissue peak activity to be able to assess the tracer influx into the tissue. In addition, tracer efflux information is also necessary to be able to estimate the tracer efflux back to plasma and the specific binding compartment. The second part ideally has to contain the 80–100 min interval to calculate SUVr, since this is the internationally conventional SUVr interval for [18F]flortaucipir [22]. So, the new scanning protocol needs to include an early part of the tracer kinetics and also at least 80–100 min post-injection (p.i.), implying that a dual-time-window protocol should be used. Scanning time can be shortened by increasing the gap of the dual-time-window. Interpolation is needed to fill this gap in the time activity curve (TAC) of the reference region to be able to perform reference tissue model–based 2

26 Chapter 2 tracer kinetic modelling. Therefore, the aim of the study is to investigate whether a shorter overall scan duration for [18F]flortaucipir PET dual-time- window scans is feasible, while retaining quantitative accuracy. METHODS Study Sample For the current project, two study cohorts were included. The first cohort consisted of ten biomarker (PET/CSF)-confirmed AD patients and ten cognitively normal controls who underwent a 130-min dynamic [18F]flortaucipir PET scan with arterial sampling (“full kinetic model cohort”). Subject characteristics have been described previously [18]. The second cohort consisted of eight subjects with AD and six cognitively normal controls that underwent two 130-min dynamic [18F]flortaucipir PET scans within a time interval of minimum 1 week, and maximum 4 weeks (“test-retest cohort”). The subject characteristics have been described previously [19]. The current study was approved by the Medical Ethics Committee of the Amsterdam University Medical Center. All subjects signed an informed consent form prior to participation. Scan Procedures T1-weighted MRI scans were acquired for all participants using a 3.0 T Philips Ingenuity Time-of-Flight PET/MR scanner (Philips medical systems, Best, the Netherlands). Isotropic structural 3D T1-weighted images were obtained using a sagittal turbo field echo sequence (1.00 mm3 isotropic voxels, repetition time = 7.9 ms, echo time = 4.5 ms, flip angle = 8°) for brain tissue segmentation. All subjects from the full kinetic model cohort underwent a 130-min dynamic [18F]flortaucipir PET scan on a Gemini TF-64 PET/CT scanner (Philips Medical Systems, Best, The Netherlands) with continuous arterial sampling after administration of 223 ± 18 MBq of [18F]flortaucipir. Details described elsewhere [17–19]. Subjects from the test-retest cohort underwent two 130-min dynamic [18F]flortaucipir PET scans on a Philips Ingenuity TF PET/CT scanner after administration of [18F]flortaucipir (237 ± 15 MBq at test and 245 ± 18 MBq at retest) as described in detail previously [19]. In short, a low-dose CT for attenuation correction was acquired, followed by a 60-min dynamic (brain) emission scan initiated simultaneously with tracer injection. After a 20-min break, a second low-dose CT was acquired before an additional dynamic emission scan during the interval 80–130 min p.i. During scanning, the head of the subjects was stabilized to reduce movement artefacts. Furthermore, subjects were positioned within the center of axial and transaxial fields of view, such that the orbitomeatal line was parallel to the detectors with the use of laser beams.

27 Shortening the scan duration For the full kinetic model cohort, continuous arterial blood sampling, using an online detection, [23] was collected during 60-min p.i. PET acquisition. Furthermore, manual arterial samples were collected at set time points (5, 10, 15, 20, 40, 60, 80, 105 and 130 min p.i.) to measure plasma metabolite fractions and plasma-to-wholeblood ratios. Using the aforementioned information, the continuous online blood sampler data was calibrated and corrected for metabolites, plasma-to-whole-blood ratios and delay, providing a metabolite-corrected arterial plasma input function. In addition, whole-blood input function was obtained for blood volume correction. Image Processing PET scans were reconstructed with a matrix size of 128 × 128 × 90 and a final voxel size of 2 × 2 × 2 mm3. All standard corrections were applied. During processing of the PET scans, first part and second part of the scan were checked for motion, separately. Thereafter, both the PET scan sessions were co-registered into a single dataset of 29 frames (1 × 15, 3 × 5, 3 × 10, 4 × 60, 2 × 150, 2 × 300, 4 × 600 and 10 × 300 s) using Vinci software (Max Plank Institute, Cologne, Germany). The last 10 frames belonged to the second PET session. Structural 3D T1-weighted MRI images were co-registered to the PET images also using Vinci software (Max Plank Institute, Cologne, Germany). The Hammers template [24], which is incorporated in PVElab [25], was used to delineate regions of interest (ROIs) on the co-registered MR scan and superimposed onto the dynamic PET scan to obtain regional time activity curves (TACs). All 68 cortical and subcortical regions from the Hammer template were included. Regional TACs extracted from the PET scans were analyzed using a reversible 2-tissue compartment model with blood volume correction (2T4k_VB) and simplified reference tissue model (SRTM) [26]. Receptor parametric mapping (RPM) [27] and standardized uptake value ratios (SUVr) were used to obtain parametric images. Cerebellum grey matter (obtained from PVElab) was used as the reference region. Shortening the Second Part of the Scan (80–130 Min P.I.) In these analyses, the first part of the scan remained 0–60 min p.i. The second part of the scan was shortened; three shorter time intervals were explored: 80–120 min, 80–110 min and 80–100 min. For each subject, shortened PET scans were acquired by removing 2 to 6 frames to reach the specified scan intervals. Reference region TACs were extracted from these shortened PET scans to estimate kinetic parameters. BPND and R1 values were estimated using RPM from the three different scan durations (0–60/80–100, 0–60/80–110 and 0–60/80–120 min). RPM-derived regional BPND and R1 values were compared to the corresponding non-linear regression (NLR)-based SRTM-derived BPND and R1, and plasma input–derived distribution volume 2

28 Chapter 2 ratio (DVR) values from the original scan duration (0–60/80–130 min). The optimal shortened time interval for the second part was used and fixed during subsequent evaluation of scan shortening of the first part of the PET scan. Shortening the First Part of the Scan (0–60 Min P.I.) For shortening the first part of the scan, three time intervals were explored: 0–50, 0–40 and 0–30 min p.i, all in combination with 80–100 min scan interval for the second part of the imaging protocol. For each subject, the corresponding frames were removed to obtain the PET scans with these specified time intervals. The original scan duration had a gap of only 20 min; the gap in the reference region was interpolated by using cubic interpolation. The larger gap (> 20 min) in the new dual-time-window protocol results in more missing data points in the reference TAC for which proper interpolation is required. Therefore, four different interpolation methods were assessed: population-based plasma input function in combination with a reversible two-tissue compartmental model with blood volume correction (POP-IP_2T4k_VB) to fit the reference tissue TAC, standard cubic interpolation, linear interpolation, and interpolation based on fitting an exponential function to the TAC (excluding points until peak uptake). All scripts were built in house using MATLAB (version R2017B, MathWorks, USA). The POP-IP_2T4K_VB interpolation method was based on using the population-averaged metabolite-corrected plasma input function and a reversible two-tissue compartmental model with blood volume correction (2T4k_VB). A 2T4k_VB model was used, since it was evaluated in the previous studies [28] that this model best describes the in vivo kinetics of [18F]flortaucipir. So based on the previous research, it was assumed that the cerebellum presents a 2T4k_VB kinetics and the cerebellum TAC with the gap was fitted using this model and the population-averaged metabolite-corrected plasma input function. The fit was visually examined for certainty and the gap in the cerebellum TAC was filled using the values from the fitted curve. SRTM-derived BPND and R1 estimates using the shortened scan durations and the four different interpolated reference region TACs were obtained. These regional parametric values were compared to the corresponding NLR-based reference region and plasma input–derived values obtained using the original scan duration (0–60/80–130 min). BPND and R1 parametric images were acquired for the optimally shortened scans with interpolated reference region (using the optimal interpolation technique(s)). Regional parametric values were extracted from these parametric images and were compared to corresponding values derived using plasma input–based and reference tissue–based NLR and RPM from the original scan duration (0–60/80–130 min).

29 Shortening the scan duration SUVr using the interval 80–100 min (SUVr(80−100 min)) was also evaluated. Regional SUVr values obtained from this time interval were compared with the respective quantitative parameters (DVR, SRTM BPND and RPM BPND) estimated using the original scan duration (0–60/80–130 min). Test-Retest Repeatability Analysis For the test-retest repeatability (TRT) analysis, the test-retest cohort was used. The TRT of RPM BPND and R1 values derived from the optimal shortened scan duration were compared to the test-retest repeatability of RPM BPND obtained using the original scan duration (0–60/80–130 min). In addition, TRT for SUVr(80–100) was also assessed. The TRT was calculated using Eq. 1. Statistical Analysis Linear regression fitting and correlation coefficients (r2) were used to compare BP ND and R1 for the shortened scan durations and SUVr(80−100 min) against corresponding parametric values for the original scan duration (0–60/80–130 min) derived from plasma input–based and reference tissue–based NLR and RPM. Furthermore, Bland-Altman plots were used to assess and illustrate TRT performance. RESULTS Shortening the Second Part of the Scan (80–130 Min P.I.) The RPM BPND values obtained from the three shortened scan durations (0–60/80– 120, 0–60/80–110 and 0–60/80–100 min) provided excellent correlations with plasma input DVR-1, SRTM BPND and RPM BPND obtained using the original acquisition time window (0–60/80–130) (Table 1; all r2 > 0.93). Reduction of the time interval of the second part to 100 min had negligible effects on the RPM BPND estimation: correspondence to DVR-1 (HC: r2 = 0.94, slope = 0.95; AD: r2 = 0.94, slope = 0.92), SRTM BP ND (HC: r2 = 0.98, slope = 1.05; AD: r2 = 0.96, slope = 0.85) and RPM BP ND (HC: r2 = 0.98, slope = 1.04; AD: r2 = 0.99, slope = 0.94). Comparison of regional SRTM BP ND values obtained using the shorter time interval (0–60/80–100) to plasma input DVR-1 and SRTM BPND obtained with the original scan duration (0–60/80–130) are presented in Supplementary Table 1. The RPM R1 values obtained from the three shortened scan durations (0–60/80– 120, 0–60/80–110 and 0–60/80–100) also provided excellent correlations with SRTM R1 and RPM R1 estimated using the original acquisition time window (0–60/80–130) (Supplementary Table 2). 2

30 Chapter 2 Table 1. RPM BPND obtained using shorter time intervals compared to plasma input DVR-1, SRTM BPND and RPM BPND obtained with the original scan duration. DVR-1 (0-60/80-130) SRTM BPND (0-60/80-130) RPM BPND (0-60/80-130) HC AD HC AD HC AD r2 Slope r2 Slope r2 Slope r2 Slope r2 Slope r2 Slope RPM BPND (0-60/80-120) 0.95 0.91 0.95 0.96 0.99 1.02 0.97 0.90 1.00 1.01 1.00 0.99 RPM BPND (0-60/80-110) 0.95 0.92 0.95 0.94 0.99 1.03 0.97 0.88 0.99 1.02 1.00 0.97 RPM BPND (0-60/80-100) 0.94 0.95 0.94 0.92 0.98 1.05 0.96 0.85 0.98 1.04 0.99 0.94 The correlation and slope for the original scan duration between RPM BPND and DVR-1 was r2= 0.95 slope= 0.91 for HC and, r2= 0.96 slope= 0.98 for AD. The correspondence between original RPM BPND and SRTM BPND was r2= 1.00 slope= 1.01 for HC and r2= 0.97 slope= 0.91 for AD.

31 Shortening the scan duration Figure 1. Interpolation of the gap in reference region TAC (30 to 80 min) with different interpolation methods. Shortening the First Part of the Scan (0–60 Min P.I.) In Figure 1, the different interpolations of a typical reference TAC for the shortest dual-time-window (0–30/80–100 min) assessed in this study are presented. For all shortened scan durations, SRTM BPND using the reference TACs interpolated with either POP-IP_2T4k_VB or cubic interpolation methods had the best correspondence with plasma input DVR-1 and SRTM BPND (r2 > 0.90, Table 2) obtained with the original scan duration. Reduction of the time interval of the first part of the scan to 30 min and using POPIP_2T4k_VB for reference region interpolation had negligible effects on the quantitative accuracy of the estimated kinetic parameters with respect to that estimated using the original scan duration: DVR-1 (HC: r2 = 0.93, slope = 0.94; AD: r2 = 0.92, slope = 0.97) and SRTM BP ND (HC: r2 = 0.96, slope = 1.02; AD: r2 = 0.98, slope = 0.89). SRTM BPND values obtained with 0–30/80–100 min data using cubic interpolation for reference region had similar agreement with DVR-1 (HC: r2 = 0.94, slope = 0.91; AD: r2 = 0.91, slope = 0.92) and SRTM BP ND (HC: r2 = 0.96, slope = 0.98; AD: r2 = 0.98, slope = 0.85) from the original scan duration. Good correlations were observed for linear and exponential interpolation methods (r2 > 0.90, Table 2). However, these interpolation methods resulted in higher underestimation (15–25 %) of the parametric values. 2

32 Chapter 2 Table 2. Shortened time intervals interpolated using four different methods are compared with plasma input DVR-1 and SRTM BPND obtained with the original scan duration DVR-1 (0-60/80-130) SRTM BPND (0-60/80-130) HC AD HC AD r2 Slope r2 Slope r2 Slope r2 Slope POP–IP 2T4k_VB SRTM BPND (0-50/80-100) 0.96 0.96 0.91 1.01 0.98 1.04 0.99 0.93 SRTM BPND (0-40/80-100) 0.95 0.96 0.91 0.99 0.98 1.03 0.99 0.91 SRTM BPND (0-30/80-100) 0.93 0.94 0.92 0.97 0.96 1.02 0.98 0.89 Cubic SRTM BPND (0-50/80-100) 0.96 0.94 0.91 1.01 0.99 1.02 0.99 0.93 SRTM BPND (0-40/80-100) 0.96 0.95 0.91 0.98 0.99 1.02 0.99 0.90 SRTM BPND (0-30/80-100) 0.94 0.91 0.91 0.92 0.97 0.98 0.98 0.85 Linear SRTM BPND (0-50/80-100) 0.96 0.93 0.91 0.98 0.99 1.02 0.99 0.90 SRTM BPND (0-40/80-100) 0.96 0.95 0.91 0.92 0.98 1.03 0.99 0.84 SRTM BPND (0-30/80-100) 0.94 0.91 0.91 0.84 0.96 0.98 0.98 0.76 Exponential SRTM BPND (0-50/80-100) 0.96 0.94 0.91 1.01 0.99 1.02 0.99 0.93 SRTM BPND (0-40/80-100) 0.96 0.96 0.90 0.97 0.98 1.04 0.98 0.90 SRTM BPND (0-30/80-100) 0.94 0.92 0.90 0.92 0.96 0.99 0.97 0.85 The correspondence of SRTM BPND with DVR-1 for the original scan duration was r2 = 0.96, slope=0.90 for HC and r2=0.93, slope=1.09 for AD subjects.

33 Shortening the scan duration Figure 2. Comparison of SRTM BPND estimated using shortened time intervals a) 0-50/80-100 min, b) 0-40/80-100 min, c) 0-30/80-100 min and POP-IP_2T4k_VB interpolation for reference region against SRTM BPND obtained from the original scan duration (0-60/80-130min). LOI = line of identity. 2

34 Chapter 2 Figure 2 presents the correspondence of SRTM BPND values obtained with the shortened scan durations using POP-IP_2T4k_VB for reference region interpolation against SRTM BPND values estimated from the original scan duration. The bias increased as the first part was shortened. An underestimation of 9 % was observed for SRTM BPND values with the shortened scan duration for both groups (0–30/80–100 min) with respect to that obtained with original scan duration. SRTM R1 values derived from the shortened scan duration (0–30/80–100 min) showed excellent correlations with SRTM and RPM R1 values obtained with the original scan duration for HC and AD patients for each interpolation method (r2 > 0.95, Supplementary Table 3).

35 Shortening the scan duration Table 3. Comparison of RPM BPND obtained with the shortest scanning interval (0-30/80-100) and SUVr(80-100min) to plasma-input DVR-1, SRTM BPND and RPM BPND derived from the original scan duration (0-60/80-130). DVR-1 (0-60/80-130) SRTM BPND (0-60/80-130) RPM BPND (0-60/80-130) HC AD HC AD HC AD r2 Slope r2 Slope r2 Slope r2 Slope r2 Slope r2 Slope POP-IP 2T4k_VB RPM BPND (0-30/80-100) 0.94 0.94 0.95 0.93 0.98 1.04 0.97 0.84 0.98 1.03 0.99 0.95 Cubic RPM BPND (0-30/80-100) 0.94 0.93 0.94 0.90 0.98 1.03 0.97 0.81 0.98 1.02 0.99 0.92 SUVr (80-100) 0.80 0.95 0.92 1.10 0.90 1.15 0.93 0.98 0.87 1.11 0.96 1.13 The correlation and slope for the original scan duration between RPM BPND and DVR-1 was r2= 0.95 slope= 0.91 for HC and, r2= 0.96 slope= 0.98 for AD. The correspondence between original RPM BPND and SRTM BPND was r2= 1.00 slope= 1.01 for HC and r2= 0.98 slope= 0.88 for AD. 2

36 Chapter 2 An example of the RPM BPND images for the original scan duration and shortened scan duration (0–30/80–100) is illustrated in Figure 3a. Comparison of RPM BPND obtained from the shortest scan duration (0–30/80–100 min) using POP-IP_2T4k_VB for reference region interpolation against RPM BPND obtained with the original scan duration is shown in Figure 3b and Supplementary Figure 1. RPM BPND obtained with the shortest scan duration (0–30/80–100 min) using either POP-IP_2T4k_VB or cubic methods for reference region interpolation and SUVr80-100 min compared to original DVR-1, SRTM BPND and RPM BPND values are shown in Table 3 for HC and AD patients separately. The same comparisons were made for RPM R1 and are illustrated in Supplementary Table 4.

37 Shortening the scan duration Figure 3. a) An example of the BPND images for a representative AD patient are displayed for the original scan (0-60/80-130 min) duration and shortened scan duration (0-30/80-100 min using POP-IP_2T4k_VB interpolation) along with the corresponding MR. b) Comparison of BPND obtained from the shortened scan duration (0-30/80-100 min using POP-IP_2T4k_VB interpolation) against BPND obtained with the original scan duration (0-60/80-130 min). c) Bland-Altman plot of the original test-retest differences for RPM DVR (BPND+1) values. d) Bland-Altman plot of the test-retest differences for RPM DVR (BPND +1) values using shortened scan duration (0-30/80-100 min) and POP-IP_2T4k_VB method for reference region interpolation. 2

38 Chapter 2 Table 4. RPM BPND, RPM R1 and SUVr values obtained from the test scan are compared to corresponding values obtained from the retest scan for the original scan duration (0-60/80-130), and for the shortened scan duration (0-30/80-100) interpolated with cubic or POP-IP_2T4k_VB interpolation method. HC AD r2 Slope %TRT r2 Slope %TRT RPM BPND (0-60/80-130) 0.91 0.95 -1 ± 4 0.98 1.0 0 ± 4 Cubic RPM BPND (0-30/80-100) 0.90 0.96 -1 ± 4 0.98 0.98 0 ± 5 POP-IP_2T4k_VB RPM BPND (0-30/80-100) 0.90 0.96 -1 ± 4 0.97 1.0 0 ± 5 RPM R1 (0-60/80-130) 0.86 0.95 0 ± 6 0.94 0.96 0 ± 6 Cubic RPM R1 (0-30/80-100) 0.86 0.94 0 ± 6 0.94 0.96 0 ± 6 POP-IP_2T4k_VB RPM R1 (0-30/80-100) 0.86 0.95 0 ± 6 0.94 0.96 0 ± 6 SUVr (80-100) 0.86 0.97 -1 ± 5 0.96 1.05 -1 ± 7 Test-Retest Repeatability The RPM BPND values estimated for both test and retest scans using the shortest assessed scan duration (0–30/80–100) and POP-IP_2T4k_VB or cubic interpolation for the reference region correlated well (Table 4). The test-retest parametric correlations using the short scan duration and interpolated reference TAC (POP-IP_2T4k_VB and cubic interpolation methods) were similar to the correlations using the original scan duration. The TRT across all Hammers’ regions of interest for RPM BPND was − 1 %± 4 for HC and 0 %± 4 for AD patients using the original dual-time-window acquisition (Table 4). Furthermore, TRT for R1 was 0 %± 6 for both groups using the original dual-time-window acquisition. The TRT for BPND remained the same for HC when using the shortest dual-time-window 0–30/80–100 min with either of the interpolations methods (POP-IP_2T4k_VB or cubic interpolation). For AD patients, the TRT changed to 0 %± 5 when using the shortest dual-time-window 0–30/80–100 min with POP-IP_2T4k_VB or cubic interpolation. The TRT for R1 remained similar as with the original data (0 %± 6) for the shortest dual-time-window (0–30/80–100 min) both with POP-IP_2T4k_VB and cubic interpolation for both groups. TRT for SUVr80-100 min was − 1 %± 5 for HC and −1 %± 6 for AD patients (Table 4). Bland-Altman plots for the RPM BPND+1 illustrating TRT for all Hammers’ regions obtained with the original scan duration and the shortest scan duration with POP-IP_2T4k_VB reference region interpolation are presented in Figure 3c and d.

39 Shortening the scan duration DISCUSSION The current study demonstrated that for [18F]flortaucipir expanding the break in the dual-time-window protocol with just a 50-min overall scanning time (early interval of 0–30 min, than a coffee break, followed by a late interval of 80–100 min) had minimal effect on the quantitative accuracy. The optimal shortened dual-time-window protocol (0–30/80–100 min) allows sufficiently accurate estimation of BPND while reducing patient burden and enables interleaved scanning, where other patients could use the camera during breaks within the scan period. An excellent correlation was observed between the shortened acquisition protocol (0–30/80–100 min) and the original dual-time-window (0–60/80–130 min) protocol. Four different interpolation methods were used to interpolate the missing data between the two time windows for the reference region TAC (cerebellum grey matter). According to our results, POP-IP_2T4k_VB interpolation, which uses a population-averaged plasma input function, showed a good correspondence of the estimated kinetic parameters to that obtained from the original scan protocol, and the lowest under/over-estimation(s) compared to other interpolation methods. Heeman et al. [29] showed that POP-IP_2T4k_VB interpolation method also works well to interpolate the missing data points in a dual-time-window protocol for [18F]flutemetamol and [18F]florbetaben. They concluded that the introduction of a gap with a maximum of 60 min in a dual-time-window protocol (early interval of 0–30 min followed by a late interval of 90–110 min) does not affect quantitative accuracy for [18F]flutemetamol and [18F]florbetaben. As such, POP-IP_2T4k_V B interpolation does not only work well for [18F]flortaucipir but also for [18F]flutemetamol and [18F]florbetaben, possibly because the model describes the in vivo kinetics of the tracers best and is therefore ideal for estimating the missing reference region data points accurately. The correlations for all interpolation methods were comparable (Table 2). This was not expected, since linear and exponential interpolation did not follow the course of tracer as can be observed in Figure 1. A possible explanation could be that the gap between the dual-time-windows was not substantially large enough to see significant differences in correlations between the interpolation methods. However, a substantial underestimation (at times up to 20 % or even more) was observed in AD patients for the shortened SRTM BPND values obtained with linear and exponential interpolation methods when compared to plasma input DVR-1 and SRTM BPND obtained with the original scan duration. This indicates that these interpolation methods are not suitable for quantitatively accurate kinetic parameter estimation for [18F]flortaucipir. For SRTM BP ND values obtained with POP-IP _2T4k_VB interpolation, the biases remained within~ 10 % for the same comparisons 2

RkJQdWJsaXNoZXIy MjY0ODMw