Introduction
Alzheimer’s disease (AD) is the most common neurodegenerative disorder and the most common cause of dementia in the elderly, which affects 47 million patients worldwide with steadily increasing numbers [
1]. The two characteristic neuropathological changes observed in AD are the deposition of extracellular amyloid senile plaques and the presence of intracellular neurofibrillary tangles (NFTs) [
2,
3]. The amyloid cascade hypothesis has been proposed and the deposition of amyloid beta (Aβ) protein is the first step of Alzheimer’s pathology including NFTs [
4]. Therefore, the deposition of Aβ protein has been the main target of in vivo diagnostic imaging tool of AD. Several imaging tracers, especially for positron emission tomography (PET), has been developed and reported to evaluate amyloid deposition, such as
11C-Pittsburgh compound B (PiB) [
5],
11C-BF227 [
6],
18F-AZD4694 [
7],
18F-FACT [
8],
18F-BAY-949172 (
18F-florbetaben) [
9],
18F-AV-45 (
18F-florbetapir) [
10], and
18F-GE067 (
18F-Flutemetamol) [
11]. PiB, the first amyloid imaging PET tracer, has been reported with successful results and used widely as a research tool [
12]. However, the short half-life of labelled
11C (20 min) limits the clinical utility of
11C-PiB as a diagnostic tracer. Therefore, several
18F-labelled beta amyloid tracers have been developed for commercial utility because
18F with half-life 110 min has been recognized as commercially available radioactive tracer for clinical practice in terms of cost, supply and shipping.
Recently, we developed a benzofuran derivative for the imaging of Aβ protein, 5-(5-(2-(2-(2-
18F-fluoroethoxy)ethoxy)ethoxy)benzofuran-2-yl)-
N-methylpyridin-2-amine (
18F-FPYBF-2) [
13]. This new fluorinated benzofuran derivative, which is like
18F-AZD4694 but has a fluoropolyethylene glycol side chain, is a promising PET probe for cerebral Aβ plaques imaging, and the specific labeling of Aβ plaques was observed in autoradiographic sections of autopsied AD brain. It should be noted that
18F-FPYBF-2 has a stable chemical structure which does not photodegrade. However, there has been no report evaluating the utility of
18F-FPYBF-2 as a PET tracer in in vivo human study.
The aim of this study was to assess the feasibility of 18F-FPYBF-2 as an amyloid imaging PET tracer in a first clinical study with healthy volunteers and in comparative dual tracer study using 11C-PiB, and to evaluate the clinical usefulness of 18F-FPYBF-2 PET/CT in the diagnosis of AD.
Discussion
Our first clinical study clearly indicated that 18F-FPYBF-2 is a safe and stable amyloid PET tracer with longer half-life with F-18 and is comparable to 11C-PiB in the detectability of amyloid deposition with high linear correlation, and that 18F-FPYBF-2 PET/CT is a useful and reliable diagnostic tool for the evaluation of AD. Although 18F-FPYBF-2 is a “late” amyloid PET tracer after the appearance of several tracers in clinical practice with comparable diagnostic ability, we would like to show the potential of 18F-FPYBF-2 as diagnostic abilities as an amyloid imaging tracer and expand the utilization of this tracer further in various fields of research and clinical practice in the following sentences.
The diagnostic abilities of
18F-FPYBF-2 as an amyloid imaging tracer are satisfactory and comparable to the other amyloid PET tracers (Figs.
1,
3). Figure
3 showed that differential diagnosis between AD patients and healthy volunteers was achieved using the qualitative analysis of Mean Cortical Index of SUVR, and that the threshold of Mean Cortical Index was about 1.2. Figure
4 left also showed that Mean Cortical Index of 1.2 in
18F-FPYBF-2 closely corresponded to Mean Cortical Index of 1.5 in
11C-PiB. It is known that Mean Cortical Index of 1.5 has been used as a threshold between AD patients and healthy volunteers in
11C-PiB PET/CT study [
26,
27]. Although the threshold value is different, we believe that diagnostic performance of
18F-FPYBF-2 PET/CT using quantitative analysis of Mean Cortical Index is enough in the differential diagnosis of AD from healthy volunteers. We have to admit that the threshold value of Mean Cortical Index 1.2 could not clearly separate AD and healthy volunteers. In the present study, there were several AD patients with low amyloid deposition and several persons with high amyloid deposition in healthy volunteer group. However, we believe that this is reasonable in clinical study. As we mentioned above, our clinical diagnosis of AD was not confirmed with pathology or others. In addition, it is known that about 10–30% of healthy aged persons showed high amyloid deposition [
28]. Therefore, we may say that our data of amyloid positive rate are comparable with the previously published data with other amyloid tracers.
While typical cases shown in Fig.
1 can be clearly diagnosed visually, we were under impression that visual diagnosis would not be easy in
18F-FPYBF-2 PET/CT. As shown in Fig.
4 and Table
3, Mean Cortical Index of AD patients observed at
18F-FPYBF-2 PET/CT is relatively lower than that at
11C-PiB PET/CT, which means that
18F-FPYBF-2 has a narrower dynamic range and lower lesion-to-normal contrast than
11C-PiB both in SUVR and in images. It may be said that the impact of
18F-FPYBF-2 as a PET imaging tool of AD is not so attractive in a visual sense compared to that of
11C-PiB.
Several amyloid PET tracers have been developed and comparative study between
11C-PiB and each PET tracer was performed so far [
29‐
33]. While some of the tracers also showed the narrower dynamic range than that of
11C-PiB, most of these reports revealed that diagnostic abilities of these amyloid PET tracers were similar and identical to that of
11C-PiB. In these PET tracers, its own specific method for visual and quantitative diagnosis is proposed in each tracer, including the indication of color or black and white tones for image display, the indication of analyzed area of brain, etc. We have to admit that we could not establish the most appropriate specific visual diagnostic method for
18F-FPYBF-2 PET/CT, so far. For the establishment of visual diagnosis, further evaluation of appropriate color scaling with comprehensive interpretation or new diagnostic method with regional area analysis or others would be needed.
It is particularly worth noting that evaluation of brain amyloid deposition in healthy volunteers was performed in broad spectrum of age range from 24 to 79 in age. Our data showed that average Mean Cortical Index of healthy volunteers (20–39, 40–49, 50–59, 60–69, 70–79 years old) were almost similar, except for the difference between the group 20–39 years old and the group 70–79 years old (
p < 0.05) (Supplement Fig. 1). A slight upward extension of the distribution was due to appearance of high SUVR subjects in the older age range. This is reasonable because it is known that, as the age increases, some normal subjects present amyloid deposition while others remain amyloid negative. Our data of 8 healthy young volunteers in the age of 20–39 can be used as a control group for the evaluation of premature senility, which is often observed in patients with Down syndrome. It is known that adults with Down syndrome even at its younger age are at a very high risk of developing early onset AD due to trisomy of chromosome 21 [
34]. We are planning to have further research using
18F-FPYBF-2 PET/CT as a diagnostic tool for the evaluation of early onset AD in Down syndrome in near future.
The limitation of the present study should be addressed. First of all, three-dimensional magnetic resonance (MR) images were not available in the study. Because the predefined AAL ROIs were masked with the gray matter of averaged standard T1-weighted structural template image in all subjects, the SUVR values in the study were influenced by inter-subject variability of gray matter volumes and the partial volume effect. Furthermore, the spatial normalization in the study was based on amyloid PET images by the amyloid PET template. Although we visually confirmed that inversely transformed AAL ROIs to an individual space corresponded to PET images in each subject, the accuracy of spatial normalization might be lower compared with MR-based normalization. The more sophisticated methods of spatial normalization only using PET images would improve both the accuracy of spatial normalization and quantification [
35‐
37]. Second, the clinical diagnosis of AD patients and other related disease was performed before PET study by board certified physicians in a comprehensive diagnosis using Japanese clinical diagnostic guideline. Therefore, there was no confirmation of the diagnosis by pathological examination or autopsy, or biomarkers in cerebrospinal fluids, such as Aβ40, Aβ42 and phosphorylated Tau (pTau). Third, in the present study, 16 patients of Mild Cognitive Impairment (MCI) showed inconclusive results of Mean Cortical Index. The results of Mean Cortical Index of SUVR in MCI patients showed relatively high and could not be clearly distinguished with those of AD patients. However, we believe that this result was reasonable. MCI was just a diagnosis at the time before PET study. It is possible that within these MCI patients there must be several patients included who will develop AD in future. Further follow-up study would be needed to clarify the outcome of the patients with high amyloid deposition.
In conclusions, our first clinical study showed that 18F-FPYBF-2 is a safe and stable amyloid PET tracer with longer half-life with F-18 and its diagnostic ability is comparable to 11C-PiB. In addition, it can be said that 18F-FPYBF-2 PET/CT is a useful and reliable diagnostic tool for the evaluation of AD by the quantitative analysis using Mean Cortical Index of SUVR, which could clearly distinguish Alzheimer disease patients by threshold of 1.2.