Impact of partial volume effect correction on cerebral β-amyloid imaging in APP-Swe mice using [18F]-florbetaben PET☆
Graphical abstract
Introduction
Alzheimer's disease (AD) is imposing a growing socioeconomic burden due to the aging population demographics, especially in industrialized nations (Ziegler-Graham et al., 2008). Present treatments are symptomatic, but some promising avenues for disease-modifying treatment are in different phases of development and approval (Singh et al., 2012). In general animal models represent an important tool for preclinical testing of disease progression and novel medications in conjunction with molecular imaging (Virdee et al., 2012). A number of rodent AD-models are characterized by cerebral overexpression of ß-amyloid and progressive formation of cortical ß-plaques, emulating this aspect of AD pathology (Teipel et al., 2011). The prospect of monitoring disease progression in individual animals has given rise to a number of rodent positron emission tomography (PET) studies in conjunction with [11C]- or [18F]-labeled ß-amyloid-tracers (Klunk et al., 2005, Maeda et al., 2007, Manook et al., 2012, Poisnel et al., 2012, Rojas et al., 2013, Snellman et al., 2012, Toyama et al., 2005). Quantitative analysis of PET data from rodent brain is necessarily compromised by partial volume effects (PVE), which result in degradation of signals from target sources of the same scale as the spatial resolution of the tomograph (Kuntner et al., 2009). However, there have been few reports of small animal PET–PVE, despite its great significance for PET quantification (Erlandsson et al., 2012). In the majority of previous human studies, PVE correction (PVEC) of brain signals has been based on the Müller–Gärtner Method (MGM) and its modifications (Thomas et al., 2011) which depend on voxel-based geometric transfer matrix (GTM) models first established by Videen and coworkers (Videen et al., 1988).
We previously investigated the progression of β-amyloid deposition in brain of AD-model mice over-expressing a human mutation of the amyloid-precursor protein (APP-Swe), in a longitudinal PET study with the novel β-amyloid tracer [18F]-florbetaben (Rominger et al., 2013). We reported a notable increase in β-amyloid PET signal from the brain of APP-Swe mice with age, which significantly correlated with histochemical findings post mortem. However, there were also distinct discrepancies between cortex-to-cerebellum standard uptake value ratios (SUVR) measured by PET in vivo and the more pronounced accumulation measured ex vivo. Similar discrepancies have been described in a study with the alternate β-amyloid tracer [18F]-AV-45 (Poisnel et al., 2012), and are likely attributable to PVE.
With this background, we aimed to establish the magnitude of PVE on quantitation of β-amyloid burden in the mouse brain with [18F]-florbetaben, and to implement and validate PVEC for our AD mouse model. We predicted that PVE-correction of PET recordings should enhance the precision of the method, this affording more sensitive detection of planned disease-modifying treatments and to reductions in required sample sizes.
Section snippets
Estimation of spatial resolution
Indwelling venous cannulae (Vasofix® Safety; B. Braun Melsungen AG, Germany) of three different internal diameters (0.4, 0.6 and 0.8 mm) were cut into 20 mm lengths, filled with a solution containing eosin and [18F] (61.56 MBq/cc), and sealed at both ends with superglue. Two sources of each diameter were placed in a cylindrical phantom chamber (volume 41.2 ml), and held in place with modeling clay at an interval of 8 mm (Fig. 1A). The phantom chamber was then filled with water containing [18F] (0.07
Estimation of spatial resolution
Mean FWHM was identical for all three internal diameters of [18F]-filled cannula sizes (F = 0.034; p = n. s.), and indicated an isotropic spatial resolution of 1.72 (± 0.08) mm (Fig. 1C).
Chessboard experiment
Cross-calibrated standard radioactivity concentrations were 225.3, 182.7, 93.6, 61.3, and 41.5 KBq/cc. Uncorrected data revealed RMSE [%] ranging from 14.2% to 24.3% (Fig. 4A), while corresponding PVE-corrected data showed RMSE [%] from 3.1% to 9.6%. Greatest reduction of RMSE [%] with PVEC was observed in wells of
Discussion
We first performed phantom studies to define with precision the point-spread function of our small animal PET recordings, and then undertook the first use of PVEC applied to mouse data with a ligand for β-amyloid, the novel tracer [18F]-florbetaben, which we have previously reported to reveal progressive amyloid accumulation in the brain of our APP-Swe mouse model (Rominger et al., 2013). In such studies, quantitation of PET signal arising from small brain regions of mouse brain is certain to
Conclusion
Application of VOI-based PVEC in [18F]-florbetaben PET studies of β-amyloid accumulation in aging APP-Swe mice improves quantitation of the PET signal, and power analysis predicts that fewer animals per group will be required to detect treatment effects in future longitudinal studies. The small reduction in group size may not be decisive in the planning of experiments, but the benefit of improved accuracy of PET measurements relative to gold standard findings is in itself compelling grounds for
Acknowledgments
We thank Rosel Oos, Karin Bormann-Giglmaier and Matthias Moser for the excellent technical support.
References (33)
- et al.
Issues in measuring glucose metabolism of rat brain using PET: the effect of harderian glands on the frontal lobe
Neurosci. Lett.
(1998) - et al.
Characterisation of partial volume effect and region-based correction in small animal positron emission tomography (PET) of the rat brain
NeuroImage
(2012) - et al.
PET imaging with [18F]AV-45 in an APP/PS1-21 murine model of amyloid plaque deposition
Neurobiol. Aging
(2012) - et al.
In vivo evaluation of amyloid deposition and brain glucose metabolism of 5XFAD mice using positron emission tomography
Neurobiol. Aging
(2013) - et al.
Development of Alzheimer-disease neuroimaging-biomarkers using mouse models with amyloid-precursor protein-transgene expression
Prog. Neurobiol.
(2011) - et al.
Evaluation of anesthesia effects on [18F]FDG uptake in mouse brain and heart using small animal PET
Nucl. Med. Biol.
(2004) - et al.
Applications of positron emission tomography in animal models of neurological and neuropsychiatric disorders
Neurosci. Biobehav. Rev.
(2012) - et al.
Worldwide variation in the doubling time of Alzheimer's disease incidence rates
Alzheimers Dement.
(2008) - et al.
Quantitative accuracy of MAP reconstruction for dynamic PET imaging in small animals
Med. Phys.
(2012) - et al.
A review of partial volume correction techniques for emission tomography and their applications in neurology, cardiology and oncology
Phys. Med. Biol.
(2012)
Imaging Abeta plaques in living transgenic mice with multiphoton microscopy and methoxy-X04, a systemically administered Congo red derivative
J. Neuropathol. Exp. Neurol.
Binding of the positron emission tomography tracer Pittsburgh compound-B reflects the amount of amyloid-beta in Alzheimer's disease brain but not in transgenic mouse brain
J. Neurosci.
Limitations of small animal PET imaging with [18F]FDDNP and FDG for quantitative studies in a transgenic mouse model of Alzheimer's disease
Mol. Imaging Biol.
Longitudinal, quantitative assessment of amyloid, neuroinflammation, and anti-amyloid treatment in a living mouse model of Alzheimer's disease enabled by positron emission tomography
J. Neurosci.
Quantification accuracy and partial volume effect in dependence of the attenuation correction of a state-of-the-art small animal PET scanner
Phys. Med. Biol.
Small-animal PET imaging of amyloid-beta plaques with [11C]PiB and its multi-modal validation in an APP/PS1 mouse model of Alzheimer's disease
PLoS One
Cited by (24)
Data on specificity of [<sup>18</sup>F]GE180 uptake for TSPO expression in rodent brain and myocardium
2018, Data in BriefCitation Excerpt :Five mice received an injection of 14.7 ± 2.2 MBq [18F]GE180 into the tail vein (150 µl saline) Seven mice received an injection of excess non-radioactive GE180 (1000:1) prior to injection of 14.5 ± 2.1 MBq [18F]GE180 (in 150 µl saline) to a tail vein. Brain and myocardium tissues were quickly removed, and radioactivity concentration was measured in a gamma counter (Cobra Quantum 5002, Packard) with decay-correction to time of tracer injection [5]. A subset of mice (N = 5 with blocking and N = 3 without blocking) was placed in the tomograph (Siemens Inveon DPET) immediately after injection of [18F]GE180, whereupon we began a 90 min dynamic emission recording as described previously [6].
Applicability of [<sup>11</sup>C]PIB micro-PET imaging for in vivo follow-up of anti-amyloid treatment effects in APP23 mouse model
2017, Neurobiology of AgingCitation Excerpt :For AD research, a positive amyloid PET scan obtained with Aβ-binding tracers such as [11C]Pittsburgh compound B ([11C]PIB) is currently recognized as a valid biomarker of brain Aβ deposition (Dubois et al., 2014) and can be exploited in clinical Aβ-centric drug trials as an inclusion criterion for AD patients in addition to longitudinal monitoring of Aβ accumulation as a secondary outcome. Micro-PET imaging of Aβ pathology in transgenic (Tg) mouse models of the disease using [11C]PIB and other Aβ-tracers provides an ideal translational method for preclinical proof-of-concept studies (Zimmer et al., 2014), and to date, changes in Aβ deposition have already been successfully followed in vivo using multiple mouse models of AD (Brendel et al., 2014; Maeda et al., 2007; Manook et al., 2012; Rominger et al., 2013; Snellman et al., 2012, 2013; Zimmer et al., 2014). However, even though the potential advantages of longitudinal amyloid PET for preclinical drug development have been suggested by the authors, micro-PET imaging of Aβ has still mainly contributed to the characterization of novel PET tracers and animal models rather than Aβ-targeted drugs.
PETPVE12: an SPM toolbox for Partial Volume Effects correction in brain PET – Application to amyloid imaging with AV45-PET
2017, NeuroImageCitation Excerpt :Atlas propagation is based on inverse warping using the deformation fields obtained from spatial normalization of the subject's MRI scan (Ashburner, 2007). Previous work has shown that a trustworthy PVEc on a region-basis should not be limited to correcting spill-over effects between different GM structures of interest, but must also consider spill-over effects from WM and extra-parenchymal background signal (Brendel et al., 2014; Thomas et al., 2011). This is especially important in the case of existing amyloid binding radiotracers, which usually exhibit high unspecific WM binding signal (Matsubara et al., 2016; Schmidt et al., 2015; Villemagne et al., 2012).
Amyloid load and translocator protein 18kDa in APPswePS1-dE9 mice: A longitudinal study
2015, Neurobiology of AgingCitation Excerpt :Quantitative analysis of PET data from either mice or rats brain are severely affected by PVE. It has been shown recently that PVE improved quantification of the PET signal in cerebral imaging of mice (Brendel et al., 2014). In this study, we applied the Reblurred Van Cittert algorithm to correct for PVE (Carasso, 1999), which makes no assumption about the homogeneity of regional uptake and does not require any anatomic information.
Improved longitudinal [<sup>18</sup>F]-AV45 amyloid PET by white matter reference and VOI-based partial volume effect correction
2015, NeuroImageCitation Excerpt :They concluded that techniques accounting for multiple regions and tissue types are more appropriate for amyloid PET, especially given that significant WM uptake is a well-known attribute of [18F]-flutemetamol, [18F]-AV45 and [18F]-florbetaben. We concur with their position, and recently undertook a microPET study testing effects of VOI-based PVEC of [18F]-florbetaben uptake in brain of AD transgenic mouse brain (Brendel et al., 2014). We were able to verify by gamma counting and autoradiography ex vivo (where PVE is avoided) that distinct overcorrections occur when VOI configurations for the mouse brain lack WM and background regions; these regions were consequently included in the VOI-based PVEC approach of human brain in the current work, which yielded plausible values and changes after PVEC.
- ☆
Financial support: The study was supported by the SyNergy Cluster.