Characterization of an 18F COX-2 PET ligand for in vivo brain imaging

Alzheimer’s disease (AD) pathogenesis is associated with early neuroinflammation, which is considered to contribute to disease progression and severity.

Therefore, understanding and regulating inflammatory pathways in the central nervous system (CNS) may contribute to prevention or delay of AD.

Cyclooxygenase-2 (COX-2) is induced in response to inflammatory stimuli, and its inhibition underlies the therapeutic efficacy of many non- steroidal anti-inflammatory drugs (NSAIDs).

COX-2 expression is significantly elevated in brain in AD and this elevation correlates with the severity of brain amyloid plaque pathology.

Hence, in vivo and non-invasive monitoring of COX-2 level in the brain can track COX-2 induction during the course of AD and also examine the clinical benefits of COX-2 inhibition in AD.

A positron emission tomography (PET) ligand to quantify the level of functional COX-2 in the brain would allow direct measurement of neuroinflammation in AD, and thereby enable disease staging and therapy evaluation. However, there is no such PET ligand is currently available.

In order to establish a clinically useful PET tracer for COX-2 imaging in AD brain, we propose to evaluate highly potent [18F]- labeled COX-2 inhibitors due to the advantages associated with the 110-minute half-life. [18F]-labeled tracers allow scanning for longer duration so as to enable equilibrium scanning and facilitate robust kinetic studies leading to accurate quantification of COX-2.

Moreover, they can be transported from a synthesis hub to nearby PET centers that lack a cyclotron and aid cost-effective PET studies.

We identified MTP, a high affinity COX-2 inhibitor (IC50 = 2.2 nM; Figure 2), possessing an aromatic site for introducing [18F]-label that is less susceptible for defluorination. MTP also has adequate lipophilicity (LogP = 2.7) to passively traverse the blood-brain barrier (BBB).

We synthesized [18F]MTP and successfully demonstrated its specific binding in COX-2 positive BxPC-3 cell lines (Figure 3, 112).

In vivo PET imaging was performed in mice, induced with lipopolysaccharide (LPS), and obtained a significantly higher binding in the brain compared to control mice, with no visible skeletal uptake (Figure 4, 150).

Subsequent in vitro autoradiography of slide-mounted sections of the harvested brain, established specific binding of the tracer in LPS-induced neuroinflammation (Figures 5 & 6).

In light of the above supporting evidence demonstrating BBB permeability, higher binding in neuroinflammation, and specific binding to COX-2, [18F]MTP will be further evaluated along with two backup ligands 2 & 3 (Figure 2), using additional PET imaging in mice models of neuroinflammation as well as AD neuropathology (Aims 1 & 2).

The most qualified tracer identified will undergo test-retest dynamic PET quantitative evaluations with concomitant arterial blood sampling in male and female anesthetized monkeys to quantify binding as total distribution volume in various brain regions (Aim 3).

We believe, the proposed experiments would lead to the selection of a successful [18F]- COX-2 PET tracer, characterized for the transition to clinical studies in AD for early diagnosis, monitoring therapeutic response, and aid development of new medications.