**2.3.3 [18F]Florbetaben ([18F]BAY94-9172 or [18F]AV-1/ZK)**

[18F]Florbetaben ([18F]BAY94-9172, previously [18F]AV-1/ZK, *trans*-4-(N-methyl-amino)-4'- [2-[2-(2-[18F]fluoro-ethoxy)-ethoxy]-ethoxy]-stilbene) is a fluorinated polyethylene glycol

Diagnosis of Dementia Using Nuclear Medicine Imaging Modalities 209

is a higher frequency of ε4 allele(s) of apolipoprotein E (APOE ε4), a strong risk factor for the development of AD, in AD patients with [18F]florbetaben-positive scans than in those with [18F]florbetaben-negative scans (Barthel, et al., 2011). These results are comparable to

Barthel *et al*. reported the largest study to date, using 81 patients with probable AD and 69 healthy controls in the primary analysis (Barthel, et al., 2011). This Phase II study confirmed the diagnostic accuracy and efficacy of [18F]florbetaben in a larger cohort of subjects. Neocortical tracer uptake was significantly higher for AD patients compared with healthy controls in frontal, temporal, parietal and occipital cortices, and anterior / posterior cingulate. On a regional level, the posterior cingulate presented the greatest separation between AD patients and healthy controls and, thus, allowed for the differentiation between the two diagnostic groups. These results are consistent with previous [11C]PiB studies that have shown that tracer uptake in the posterior cingulate is a good and early indicator of probable AD (Ziolko, et al., 2006). Visual assessment of [18F]florbetaben scans have a sensitivity of 80% and specificity of 91% for distinguishing AD patients from healthy controls. Upon statistical adjustment, due to the fact that somewhat unreliable clinical diagnoses (70-90% accuracy) were used as the standard of truth, the sensitivity increased to

While Phase II clinical trial showed promising results, [18F]florbetaben imaging needs to be compared to *post-mortem* studies to validate this technique. As of November 30, 2009, [18F]florbetaben has entered Phase III FDA-approved clinical studies in the US (Bayer, Accessed 2011). Using approximately 232 individuals, this trial (Clinical Trial NCT01020838) will assess the efficacy of [18F]florbetaben PET imaging for detection of cerebral Aß against the golden standard of *post-mortem* histopathology. The trial is not expected to be completed

[18F]Florbetapir ([18F]AV-45, (E)-2-(2-(2-(2-[18F]fluoroethoxy)ethoxy)ethoxy,)-5-(4 methylaminostyryl)pyridine, Amyvid) was recently developed by Avid Radiopharmaceuticals. As a fluoropegylated stilbene derivative, [18F]Florbetapir is similar in structure to [18F]florbetaben, but the stilbene backbone has been replaced with a styrylpyridine core (Zhang, et al., 2007; Zhang, et al., 2005a). [18F]Florbetapir was chosen from a small number of 18F-labeled styrylpyridine analogs due to its optimum *in vivo* kinetics and high selectivity for Aβ plaques (Zhang, et al., 2007). Pre-clinical characterization of [18F]florbetapir demonstrated that this radiotracer has excellent binding affinity (Kd=3.72±0.30 nM) to Aβ aggregates, moderate lipophilicity, high initial brain uptake (7.33±1.54% injected dose per gram at 2 min post-injection) and rapid washout kinetics (1.88±0.14% injected dose per gram remaining in the brain 60 minutes post injection) in normal mice and primate brain (Choi, et al., 2009). Additionally, as supported by *in vitro* autoradiography of *post-mortem* human brain tissue sections and *ex vivo* autoradiography of transgenic mouse brain, [18F]florbetapir selectively labels fibrillar Aβ plaques, but not tau NFTs (Choi, et al., 2009; Choi, et al., 2011). Non-specific binding is low or non-existent. This spatial distribution of [18F]florbetapir uptake is similar to the pattern observed for

Favorably, [18F]florbetapir also exhibits fast brain kinetics comparable to that of [11C]PiB (McNamee, et al., 2009). The signal-to-noise ratio for this radiotracer asymptotes at 50-60

findings using [11C]PiB.

until April 2014.

96% while the specificity was determined to be 97%.

**2.3.4 [18F]Florbetapir ([18F]AV-45 or Amyvid)** 

[18F]florbetaben (Zhang, et al., 2005b).

(PEG) stilbene derivative that was developed by Zhang *et al*. in 2005 through Avid Radiopharmaceuticals under the name [18F]AV-1/ZK (Zhang, et al., 2005a). Due to their high binding affinities to Aβ aggregates, stilbene derivatives have been considered as potential Aβ-targeting probes for PET imaging (H. F. Kung, et al., 2001; Zhang, et al., 2005a; Zhang, et al., 2005b). Stilbene derivatives have similar structural characteristics as benzothiazoles (PiB derivatives) and thus compete for the same binding site on amyloid plaques (Zhang, et al., 2005b). [18F]Florbetaben was chosen from a series of stilbenes for its appropriate lipophilicity, high binding affinity (Ki= 6.7±0.3 nM), good safety profile, high initial uptake and rapid washout in normal mouse brain (4 minutes post-injection), and being 18F-radiolabeled (Barthel, et al., 2010; H. F. Kung, et al., 2001; Patt, et al., 2010). The typical administered dose for [18F]florbetaben is 350 MBq (9.5 mCi), and the effective dose is 5.13 mSv using this dose (whole-body effective dose conversion factor= 14.67±1.39 μSv/MBq) (O'Keefe, et al., 2009). However, [18F]florbetaben causes a much lower radiation burden than most other 18F-labeled probes, such as [18F]flutemetamol (Koole, et al., 2009). *In vivo* autoradiography of transgenic mouse brain and *in vitro* autoradiography of *post-mortem* human AD brain tissues have demonstrated specific [18F]florbetaben binding to neuritic Aβ plaques and to cerebral amyloid angiopathy (Zhang, et al., 2005b); [18F]florbetaben does not label Lewy bodies, Pick bodies, glial cytoplasmic inclusions, α-synuclein, NFTs, or other tau pathology to any appreciable extent (Zhang, et al., 2005b). Due to convincing pre-clinical data, [18F]florbetaben entered clinical trials, under the license of Bayer Schering Pharma.

Rowe *et al*. performed the first-in-man proof of mechanism study in 2008, using 15 elderly healthy controls, 15 patients with probable AD, and 5 patients with frontotemporal lobe degeneration (FTLD) (Rowe, et al., 2008). AD patients showed extensive cortical [18F]florbetaben uptake while healthy controls and FTLD patients only demonstrated nonspecific binding in white matter, but at levels comparable to those of AD patients. Visual assessment of [18F]florbetaben scans correctly differentiated AD patients from both healthy controls and FTLD patients in all but two cases, leading to a sensitivity and specificity of 100% and 90%, respectively. Results for this study are comparable with previous [11C]PiB studies, in terms of radiotracer distribution. However, the amount of tracer uptake is slightly higher for [11C]PiB. At 90-120 minutes post-injection, the mean neocortical SUVR, with the cerebellum as the reference region, for [18F]florbetaben was 57% greater in AD patients than in healthy controls; but in comparison, [11C]PiB binding is, on average, 70% higher in AD patients than in healthy controls (Rowe, et al., 2007). Nevertheless, [18F]florbetaben PET imaging was shown to be highly sensitive and specific in discriminating between AD patients, healthy controls, and FTLD patients, at a level similar to [11C]PiB.

Subsequent studies have confirmed [18F]florbetaben's high discriminatory power, low interand intrareader variability, and clinical utility (Barthel, et al., 2010). As an optimal imaging agent, [18F]florbetaben reaches maximum effect size fairly quickly (90 minutes postinjection) and maintains this contrast for up to 4 hours post-injection (Barthel, et al., 2010; Rowe, et al., 2007). This implies a certain flexibility of the imaging time window, which is advantageous in the clinical arena. [18F]Florbetaben has also been useful in better clarifying the relationship between Aβ load and other AD biomarkers. For example, regional radiotracer uptake, particularly in the lateral temporal cortex, is slightly, but significantly inversely proportional to MMSE and word-list memory scores when AD patients and healthy controls are pooled together (Barthel, et al., 2011; Rowe, et al., 2007). Moreover, there

(PEG) stilbene derivative that was developed by Zhang *et al*. in 2005 through Avid Radiopharmaceuticals under the name [18F]AV-1/ZK (Zhang, et al., 2005a). Due to their high binding affinities to Aβ aggregates, stilbene derivatives have been considered as potential Aβ-targeting probes for PET imaging (H. F. Kung, et al., 2001; Zhang, et al., 2005a; Zhang, et al., 2005b). Stilbene derivatives have similar structural characteristics as benzothiazoles (PiB derivatives) and thus compete for the same binding site on amyloid plaques (Zhang, et al., 2005b). [18F]Florbetaben was chosen from a series of stilbenes for its appropriate lipophilicity, high binding affinity (Ki= 6.7±0.3 nM), good safety profile, high initial uptake and rapid washout in normal mouse brain (4 minutes post-injection), and being 18F-radiolabeled (Barthel, et al., 2010; H. F. Kung, et al., 2001; Patt, et al., 2010). The typical administered dose for [18F]florbetaben is 350 MBq (9.5 mCi), and the effective dose is 5.13 mSv using this dose (whole-body effective dose conversion factor= 14.67±1.39 μSv/MBq) (O'Keefe, et al., 2009). However, [18F]florbetaben causes a much lower radiation burden than most other 18F-labeled probes, such as [18F]flutemetamol (Koole, et al., 2009). *In vivo* autoradiography of transgenic mouse brain and *in vitro* autoradiography of *post-mortem* human AD brain tissues have demonstrated specific [18F]florbetaben binding to neuritic Aβ plaques and to cerebral amyloid angiopathy (Zhang, et al., 2005b); [18F]florbetaben does not label Lewy bodies, Pick bodies, glial cytoplasmic inclusions, α-synuclein, NFTs, or other tau pathology to any appreciable extent (Zhang, et al., 2005b). Due to convincing pre-clinical data, [18F]florbetaben entered clinical trials, under the license of Bayer Schering Pharma. Rowe *et al*. performed the first-in-man proof of mechanism study in 2008, using 15 elderly healthy controls, 15 patients with probable AD, and 5 patients with frontotemporal lobe degeneration (FTLD) (Rowe, et al., 2008). AD patients showed extensive cortical [18F]florbetaben uptake while healthy controls and FTLD patients only demonstrated nonspecific binding in white matter, but at levels comparable to those of AD patients. Visual assessment of [18F]florbetaben scans correctly differentiated AD patients from both healthy controls and FTLD patients in all but two cases, leading to a sensitivity and specificity of 100% and 90%, respectively. Results for this study are comparable with previous [11C]PiB studies, in terms of radiotracer distribution. However, the amount of tracer uptake is slightly higher for [11C]PiB. At 90-120 minutes post-injection, the mean neocortical SUVR, with the cerebellum as the reference region, for [18F]florbetaben was 57% greater in AD patients than in healthy controls; but in comparison, [11C]PiB binding is, on average, 70% higher in AD patients than in healthy controls (Rowe, et al., 2007). Nevertheless, [18F]florbetaben PET imaging was shown to be highly sensitive and specific in discriminating between AD patients, healthy controls, and FTLD patients, at a level similar

Subsequent studies have confirmed [18F]florbetaben's high discriminatory power, low interand intrareader variability, and clinical utility (Barthel, et al., 2010). As an optimal imaging agent, [18F]florbetaben reaches maximum effect size fairly quickly (90 minutes postinjection) and maintains this contrast for up to 4 hours post-injection (Barthel, et al., 2010; Rowe, et al., 2007). This implies a certain flexibility of the imaging time window, which is advantageous in the clinical arena. [18F]Florbetaben has also been useful in better clarifying the relationship between Aβ load and other AD biomarkers. For example, regional radiotracer uptake, particularly in the lateral temporal cortex, is slightly, but significantly inversely proportional to MMSE and word-list memory scores when AD patients and healthy controls are pooled together (Barthel, et al., 2011; Rowe, et al., 2007). Moreover, there

to [11C]PiB.

is a higher frequency of ε4 allele(s) of apolipoprotein E (APOE ε4), a strong risk factor for the development of AD, in AD patients with [18F]florbetaben-positive scans than in those with [18F]florbetaben-negative scans (Barthel, et al., 2011). These results are comparable to findings using [11C]PiB.

Barthel *et al*. reported the largest study to date, using 81 patients with probable AD and 69 healthy controls in the primary analysis (Barthel, et al., 2011). This Phase II study confirmed the diagnostic accuracy and efficacy of [18F]florbetaben in a larger cohort of subjects. Neocortical tracer uptake was significantly higher for AD patients compared with healthy controls in frontal, temporal, parietal and occipital cortices, and anterior / posterior cingulate. On a regional level, the posterior cingulate presented the greatest separation between AD patients and healthy controls and, thus, allowed for the differentiation between the two diagnostic groups. These results are consistent with previous [11C]PiB studies that have shown that tracer uptake in the posterior cingulate is a good and early indicator of probable AD (Ziolko, et al., 2006). Visual assessment of [18F]florbetaben scans have a sensitivity of 80% and specificity of 91% for distinguishing AD patients from healthy controls. Upon statistical adjustment, due to the fact that somewhat unreliable clinical diagnoses (70-90% accuracy) were used as the standard of truth, the sensitivity increased to 96% while the specificity was determined to be 97%.

While Phase II clinical trial showed promising results, [18F]florbetaben imaging needs to be compared to *post-mortem* studies to validate this technique. As of November 30, 2009, [18F]florbetaben has entered Phase III FDA-approved clinical studies in the US (Bayer, Accessed 2011). Using approximately 232 individuals, this trial (Clinical Trial NCT01020838) will assess the efficacy of [18F]florbetaben PET imaging for detection of cerebral Aß against the golden standard of *post-mortem* histopathology. The trial is not expected to be completed until April 2014.
