Introduction
[F-18]-AV-1451 (Flortaucipir) is a novel positron emission tomography (PET) tracer that preferentially binds to paired helical filament (PHF)-tau containing neurofibrillary tangles (NFTs) in Alzheimer’s disease (AD) brains [
33,
51] and those that form as a function of age [
31,
33]. Recent data have also shown that [F-18]-AV-1451 binding in legacy postmortem material closely correlates with NFT Braak staging and regional tau burden [
34], suggesting that [F-18]-AV-1451 holds promise as a biomarker for the in vivo staging and quantification of tau pathology in AD. The affinity of this tracer for tau aggregates composed of straight filaments in non-AD tauopathy cases remains controversial [
31‐
33,
39,
42]. Several studies, including our own, have shown that [F-18]-AV-1451 does not bind to a significant extent to β-amyloid, α-synuclein or TDP-43-containing lesions [
31,
33,
42].
An increased in vivo [F-18]-AV-1451 retention in midbrain, basal ganglia and choroid plexus has been observed in a high percentage of elderly individuals regardless of their clinical diagnosis; including not only patients clinically diagnosed with AD [
5,
8,
18,
40] and other non-AD tauopathies [
7,
9,
11,
13,
19,
32,
38,
44,
45,
47,
49], but also patients with Parkinson’s disease (PD) and other synucleinopathies [
9,
10,
21] as well as clinically normal individuals [
5,
8,
9,
18,
26,
40,
45] whose brains are not anticipated to harbor tau pathology in those regions.
Our previous work using [F-18]-AV-1451 autoradiography in postmortem brain tissue revealed that the nearly universal midbrain uptake observed in vivo seems heavily influenced by the tracer off-target binding to neuromelanin-containing neurons in the substantia nigra (SN) [
32,
33]. The basis for increased in vivo [F-18]-AV-1451 retention frequently seen in basal ganglia and choroid plexus, however, remains unknown. To date, only a few [F-18]-AV-1451 imaging-pathological correlation studies have been conducted on either single cases or small series of autopsy-confirmed non-AD tauopathies [
27,
32,
36,
44,
46] yielding conflicting results. We have suggested that tau conformation (specifically, paired helical vs. straight tau filaments) may be critical for [F-18]-AV-1451 binding, limiting the potential usefulness of this tracer for in vivo detection of tau in many non-AD tauopathies [
32,
33]. Of note, in nearly all published autopsy-confirmed non-AD tauopathy cases imaged, the highest in vivo signal and postmortem tau pathology burden were noted in basal ganglia. However, many other regions in these cases also contained high amounts of tau aggregates at postmortem but exhibited very little or no in vivo signal. These findings suggest a potential off-target binding of this tracer within brain regions of interest in many non-AD tauopathies; in particular, off-target binding in the basal ganglia would confound possible detection of tau lesions within the basal ganglia.
Literature on [F-18]-AV-1451 PET imaging in patients clinically diagnosed with α-synucleinopathies is still scarce [
17,
20,
28]. A recent study reported increased in vivo tracer retention in patients with dementia with Lewy bodies (DLB) and cognitively impaired PD patients in inferior temporal cortex and precuneus that correlated well with severity of cognitive deficits [
17]. Another study observed that in vivo [F-18]-AV-1451 retention is significantly lower in DLB compared to AD patients, especially in the medial temporal lobe, but elevated in posterior temporoparietal and occipital cortices relative to controls [
28]. Another study showed that in vivo [F-18]-AV-1451 retention in PD patients with mild cognitive impairment is not significantly different than that of healthy controls and it does not correlate with cognitive dysfunction. Even though no imaging-pathological correlation studies have been published so far in DLB or PD patients, it is well-established the overlap of α-synuclein-containing lesions with AD pathology in many of them; something that likely accounts, at least in part, for the tracer retention observed in vivo in some of these patients [
24,
25,
41].
We have had the opportunity to study in detail the [F-18]-AV-1451 imaging-pathologic correlates in an autopsy-confirmed PD case and have used this to investigate the off-target in vivo signal observed in this patient in midbrain, basal ganglia, choroid plexus and some focal areas in the cortex. Additional legacy postmortem material containing basal ganglia, choroid plexus and parenchymal hemorrhages from 20 subjects (including controls free of pathology, AD, non-AD tauopathies, DLB, vascular dementia, and cerebral amyloid angiopathy (CAA)) were also studied for comparison purposes to better understand what [F-18]-AV-1451 in vivo positivity in those regions means.
Discussion
This is the first imaging-pathological correlation of novel PHF-tau PET tracer [F-18]-AV-1451 in an autopsy-confirmed PD case with minimal-to-none AD co-pathology. The study of this single case has been particularly informative to learn new and valuable information about the frequently observed in vivo off-target retention of this tracer in brain regions like midbrain, basal ganglia and choroid plexus, and investigate the underlying substrate/s that may be responsible for such signal in the absence of brain tau pathology. Importantly, this same PET pattern is frequently observed in elderly individuals, including those clinically normal [
5,
8,
9,
18,
26,
40,
45]. Thus, this PD case sheds light on how to correctly interpret [F-18]-AV-1451 PET in vivo images.
Our experiments using [F-18]-AV-1451 phosphor screen and high resolution autoradiography in multiple brain regions from this PD case showed that this tracer bound with strong affinity to age-related NFTs in the EC, neuromelanin-containing neurons in the substantia nigra and leptomeningeal melanocytes adjacent to the lateral ventricles, and to a lesser extent to microhemorrhages in the cortex. All these findings are consistent with our previously published observations [
32,
33]. In contrast, no detectable [F-18]-AV-1451 binding was observed in basal ganglia or choroid plexus, the two regions that displayed the highest in vivo tracer retention in this case (SUVR of 1.7 and 1.5, respectively); these data are also in agreement with our prior findings [
32]. The study of this PD case and additional brain material from 20 individuals with various neurodegenerative diagnoses have allowed us to further define the underlying substrates of in vivo [F-18]-AV-1451 retention in these two regions.
In our previous studies we observed robust off-target binding of [F-18]-AV-1451 to neuromelanin- and melanin-containing cells and alerted on the importance of carefully taking this finding into account when interpreting [F-18]-AV-1451 in vivo retention patterns [
32,
33]. Other authors have made similar observations and suggested that this off-target binding may actually be of utility to assess dopaminergic cell loss in PD patients [
21].
As noted above, elevated in vivo [F-18]-AV-1451 retention in basal ganglia has been observed in a significant proportion of elderly individuals with different clinical diagnosis, including AD [
5,
8,
18,
40] and non-AD tauopathies [
7,
10,
11,
13,
19,
27,
32,
36,
38,
44,
45,
47,
49], but also in cases without suspected underlying tau pathology like PD [
9,
20] and MSA [
10] as well as in clinically normal individuals [
5,
8,
9,
18,
26,
40,
45]. Our previous studies, including correlations in 3 non-AD tauopathy cases who underwent imaging prior to death (two PSP and a MAPT P301L mutation carrier), showed elevated in vivo retention and tau pathology in basal ganglia, but no tracer binding in this region at postmortem by autoradiography, and no significant correlation between in vivo signal and tau burden in multiple ROIs [
32]. The study is the largest series published to date on non-AD taupathies and made us conclude that tracer in vivo signal in basal ganglia in these cases was likely representing off-target retention in on-target areas for those diseases. The PD case studied here, with high in vivo retention but no tau-containing lesions or calcifications in this area, further reinforces this idea. Interestingly, several [F-18]-AV-1451 kinetic modeling studies [
1,
2,
43,
50] have suggested that this tracer has a different kinetic profile in the putamen, with a higher initial uptake and much faster clearance in this region compared to the cortex, and enhanced retention with increasing age. It has been proposed that this may be due to additional off-target binding in the putamen or a second binding site in this region with different kinetics.
To further investigate the mismatch between elevated in vivo [F-18]-AV-1451 retention in basal ganglia and lack of autoradiography signal in this region, we performed [F-18]-AV-1451 phosphor screen and high resolution autoradiography in basal ganglia sections from 12 cases with various neurodegenerative diseases (Table
1
, Fig.
4). The absence of tracer binding in this region across cases, regardless of the presence or absence of tau-containing lesions suggests that the in vivo signal in this area may be due, at least in part, to non-specific biological or technical factors unrelated to tau or non-tau substrates. However, we cannot rule out with absolute certainty that the autoradiography techniques at postmortem may remove some weak [F-18]-AV-1451 labeling from the basal ganglia.
Another brain region exhibiting potential [F-18]-AV-1451 off-target retention is the choroid plexus, a highly vascular region mostly composed of an overlying specialized epithelial layer with a stroma containing blood vessels, sometimes with focal calcifications particularly in older subjects, and small rests of meningothelial elements. Elevated in vivo tracer retention was observed in the choroid plexus in the PD case reported here but, similarly to the basal ganglia, no tau pathology could be demonstrated in this area at postmortem, and autoradiography failed to show significant tracer binding. Increased in vivo retention in the choroid plexus is a common finding in a high percentage of individuals undergoing [F-18]-AV-1451 PET scans, and especially in African-Americans (Lee CM et al., communication at the Human Amyloid Imaging conference, 2017). Of note, due to the close location of choroid plexus to medial temporal lobe structures, elevated in vivo signal in this area can potentially interfere with assessment of “true” tracer retention in the hippocampus and entorhinal cortex; thus, it is important to understand the underlying substrate of tracer’s uptake in the choroid plexus. Our autoradiography study of postmortem tissue samples, which included choroid plexus from 6 individuals, detected tracer binding in three of them corresponding to the presence in these cases of abundant leptomeningeal melanocytes (see representative cases in Fig.
5a-b). These data suggest that off-target binding to melanin contributes, at least in part, to in vivo tracer retention in choroid plexus. But the PD case reported here also reveals that in vivo signal in this region may as well be present in the absence of tau pathology or melanin, pointing to an alternative substrate. It is also possible that there is a distinct kinetic profile of the compound in this area that contributes to in vivo signal but is not captured by our autoradiographic methods.
In our PD case we also noted increased in vivo [F-18]-AV-1451 retention in focal areas of frontal and occipital cortices in the left hemisphere. Our autoradiography experiments revealed, in the limited number of sections analyzed, the presence of tracer binding to an occipital microhemorrhage. It is conceivable that our PD case may harbor additional microhemorrhages that would only be revealed by extensive brain sampling. The analysis of additional legacy postmortem material from two CAA cases harboring multiple brain hemorrhages further confirmed tracer binding to those lesions in autoradiography (Fig.
6). This is in agreement with our previously published observations indicating that the off-target binding of this tracer also includes blood products [
33]. Also, a recent publication describing 3 cases with probable CAA imaged with PET-[F-18]-AV-1451 showed that regions with microbleeds largely overlapped with those with increased [F-18]-AV-1451 in vivo retention [
29].
Conclusion
In conclusion, the imaging-pathologic correlation analysis of the first autopsy-confirmed PD patient who underwent [F-18]-AV-1451 PET scan prior to death confirms that this tracer not only binds with strong affinity to NFT tau pathology in AD, but also exhibits off-target binding to neuromelanin and melanin-containing cells and, to a lesser extent, to brain hemorrhagic lesions. These substrates likely explain, at least in part, the enhanced PET in vivo signal frequently noticed in midbrain, basal ganglia and choroid plexus regardless of the clinical diagnosis and of the presence or absence of tau-containing lesions in those regions. However, the robust off-target in vivo retention in basal ganglia and choroid plexus, in the absence of tau deposits, meningeal melanocytes or any other identifiable binding substrate by autoradiography in the PD case reported here, suggests that differential uptake and clearance profiles of this compound in these brain regions deserve to be further investigated. All together these data offer new important clues for the accurate interpretation of the patterns of [F-18]-AV-1451 retention observed by in vivo neuroimaging. Additional imaging-pathological studies on postmortem material from individuals studied by imaging methods prior to death will continue to provide insight into the implications of [F-18]-AV-1451 signals.
Acknowledgements
We are grateful to the study subjects, the MGH Gordon PET Core for providing [F-18]-AV-1451 and Dr. Peter Davies, from the Feinstein Institute for Medical Research, for kindly sharing the PHF-1 antibody.
Competing interests
Marta Marquié received research funding from the ASISA foundation, Madrid, Spain.
Eline Verwer received research funding from the Society of Nuclear Medicine and Molecular Imaging Education and Research Foundation (Cassen Postdoctoral Mentoring Award to MDN).
Sally Ji Who Kim received research funding from NIH T32 EB013180.
Cinthya Agüero received research funding from the International Health Central America Institute, San Jose, Costa Rica.
Marc D. Normandin received research funding from NIH National Institute of Neurological Disorders and Stroke (U01NS086659) and NIH National Institute of Mental Health (R01MH100350).
Stephen N. Gomperts received research funding from NIH National Institute for Neurological Disorders and Stroke (R21 NS090243), the National Parkinson’s Foundation, and the Michael J. Fox Foundation for Parkinson’s Research.
Keith A. Johnson received research funding from NIH (grants R01 EB014894, R21 AG038994, R01 AG026484, R01 AG034556, P50 AG00513421, U19 AG10483, P01 AG036694, R13 AG042201174210, R01 AG027435 and R01 AG037497) and the Alzheimer’s Association (ZEN-10-174,210 K). Keith A. Johnson has served as paid consultant for Bayer, GE Healthcare, Janssen Alzheimer’s Immunotherapy, Siemens Medical Solutions, Genzyme, Novartis, Biogen, Roche, ISIS Pharma, AZTherapy, GEHC, Lundberg, and Abbvie. He is a site co-investigator for Lilly/Avid, Janssen Immunotherapy and Pfizer.
Matthew P. Frosch received research funding from the Massachusetts Alzheimer’s Disease Research Center (NIH AG005134).
Teresa Gómez-Isla received research funding from NIH National Institute on Aging (AG005134 and AG036694).