Impact of amyloid imaging on drug development in Alzheimer's disease

Abstract

Imaging agents capable of assessing amyloid-beta (Abeta) content in vivo in the brains of Alzheimer's disease (AD) subjects likely will be important as diagnostic agents to detect Abeta plaques in the brain as well as to help test the amyloid cascade hypothesis of AD and as an aid to assess the efficacy of anti-amyloid therapeutics currently under development and in clinical trials. Positron emission tomography (PET) imaging studies of amyloid deposition in human subjects with several Abeta imaging agents are currently underway. We reported the first PET studies of the carbon 11-labeled thioflavin-T derivative Pittsburgh Compound B in 2004, and this work has subsequently been extended to include a variety of subject groups, including AD patients, mild cognitive impairment patients and healthy controls. The ability to quantify regional Abeta plaque load in the brains of living human subjects has provided a means to begin to apply this technology as a diagnostic agent to detect regional concentrations of Abeta plaques and as a surrogate marker of therapeutic efficacy in anti-amyloid drug trials.

Figure 1

(A) Post mortem fluorescent microscopy image of an 8 μm thick section from AD brain frontal cortex stained with X-34 showing amyloid plaques and neurofibrillary tangles; (B) Post mortem fluorescent microscopy image of an 8 μm thick section from an aged, non-demented control brain stained with X-34 showing lack of pathology.

Figure 2

Structures of thioflavin-T and a neutral, lipophilic derivative of thioflavin-T, PiB. The site of radiolabelling with ¹¹C is indicated with an asterisk (*).

Figure 3

PET images in transaxial and sagittal planes of parametric Logan distribution volume ratios (DVR, relative to cerebellum) of [¹¹C]PIB (370–555 MBq, 90 min scans) in a normal control subject (NC), [¹¹C]PIB-positive normal control (NC+), [¹¹C]PIB-negative mild cognitive impairment subject (MCI-), [¹¹C]PIB-positive MCI subject (MCI+), highly [¹¹C]PIB-positive MCI subject (MCI++), and a [¹¹C]PIB-positive Alzheimer’s disease (AD) subject. The small amount of signal seen in the NC- and MCI- scans in white matter areas (including brain stem) reflects slowly clearing, non-specifically bound tracer. The low level signal in the NC+ subject indicates specifically bound tracer in frontal cortex, a brain region often among the first to show [¹¹C]PIB binding in cognitively normal control subjects. The three MCI subjects show highly variable amounts of [¹¹C]PIB binding, with the MCI- subject nearly identical to the NC- subject and the MCI++ nearly identical to the AD subject.

Figure 4

Schematic of the hypothetical progression of Aβ deposition over time from the early initiation (ei) phase, to the continuously progressive (p) phase, and finally the late equilibrium (eq) phase. Subjects may experience either a long (p₁/t₁) or brief (p₂/t₂) progressive phase of Aβ deposition, which snapshots such as shown in Figure 3 can capture at a single point in time. Cognitive symptoms may not be evident until the equilibrium (eq) phase (MCI+, MCI++, and AD in Figure 3), but the cascade of pathological events that may lead to these symptoms (i.e. neurofibrillary pathology, synapse and neuron loss) may be initiated during the progressive phase (p) (figure from [94])

Figure 5

Distribution volume ratio (DVR) outcome measures of [¹¹C]PiB for individual control (n=8), MCI (n=10), and AD (n=6) subjects for posterior cingulate gyrus (PCG) and frontal cortex (FRC) regions of interest. The regional DVR outcome measures were determined using the non-invasive Logan graphical analysis method with cerebellum as the reference region over 90 min post-injection. The numbered circles represent the individual subjects, and subjects with overlapping values are placed adjacent to one another. Note that the MCI group displays DVR values ranging from control-like (<1.4) to AD-like (>2.0) with a few MCI subjects displaying intermediate DVR values (between 1.5 and 2.0). Adapted from [22].

Figure 6

Schematic of the production of Aβ (1–40) and Aβ (1–42) peptides from amyloid precursor protein (APP) via sequential β-and γ-secretase cleavages. The Aβ (1–40) and Aβ (1–42) monomers are believed to aggregate in the excellular space to form soluble Aβ species (oligomers). Some Aβ oligomers are believed to form proto-fibrils, which then form slightly soluble Aβ fibrils and plaques. Oligomers and Aβ-containing fibrils and plaques are believed to set in motion subsequent damaging processes, such as tau hyperphosphorylation and neurofibrillary tangle formation, generation of excitotoxic species, oxidative damage, neuroinflammation, enlargement of axons and dendrites with deposits of hyperphosphorylated tau filaments (dystrophic neurites), loss of synaptic junctions, and neuronal cell death.

Figure 7

Post mortem fluorescent microscopy images of X-34 staining of 8 μm thick brain sections from an AD patient treated with the anti-amyloid therapy AN-1792 (an active immunization treatment). Sections include tissues frontal cortex (A) and temporal cortex (B) showing mostly neurofibrillary tangles and other tau-containing cellular processes, but the sections are relatively free of Aβ plaques. The asterisks (*) mark “holes” in the brain were Aβ plaques may have previously existed prior to AN-1792 treatment. The white scale bar is 100 μm in length.

Figure 8

Transaxial and sagittal [¹¹C]PiB parametric images of the reference Logan distribution volume ratio (DVR) using cerebellar tissue as input in an elderly normal control subject at baseline and 2 years later. The subject shows increased, unilateral [¹¹C]PiB binding in the frontal cortex (arrows) over the 2 year time period, perhaps suggesting a future course of cognitive decline.

Figure 9

In vitro binding assays of [³H]PiB in post mortem homogenized tissues taken from the frontal cortex (Fr) and cerebellum (CB) of normal control subjects (Cntl), untreated Alzheimer’s disease subjects (AD), and the AN-1792-treated AD subject shown in Figure 7. All cerebellar tissues show low levels of specifically bound tracer in accordance with the absence of Aβ plaque pathology in the cerebellum, but the frontal cortex from the untreated AD subjects shows a high level of specifically bound tracer. Frontal cortex from the AN-1792 treated AD subject shows values similar to control subjects, indicating the absence of Aβ plaque pathology. The question is: did this AN-1792 treated subject have Aβ plaque pathology in the frontal cortex prior to treatment? Imaging studies using Aβ imaging agents such [¹¹C]PiB at baseline and after anti-amyloid treatment will be able to answer this question.

Figure 10

Schematic of the time course of Aβ deposition and the offset in time of subsequent cognitive decline. The dotted line under AD indicates the time point for application of anti-amyloid treatments in clinically apparent AD patients, while the dotted NC+ and MCI+ lines indicate the time points for application of these treatments when cognitive performance is relatively intact, but Aβ deposition is apparent. Treatments with anti-amyloid therapies at these earlier time points could halt the downstream damage caused by excessive cerebral Aβ loads (see Figures 4 and 6).