First-in-Human Study of 18F-Labeled PET Tracer for Glutamate AMPA Receptor [18F]K-40: A Derivative of [11C]K-2

J Nucl Med. 2025 May 8:jnumed.124.269405. doi: 10.2967/jnumed.124.269405. Online ahead of print.

ABSTRACT

Although the alteration of glutamate α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor (AMPAR) distribution is believed to underlie physiologic and pathologic neuronal function, there has been no modality to evaluate AMPARs in a living human. [¹¹C]K-2, the PET tracer we previously developed, is the first and only technology, to the best of our knowledge, to visualize AMPAR densities in the living human brain. Despite its favorable kinetics as a PET tracer, the short half-life of ¹¹C limits the potential of [¹¹C]K-2. We recently developed an ¹⁸F-labeled PET tracer, [¹⁸F]K-40, which demonstrated AMPAR-specific binding properties and brain distribution similar to that of [¹¹C]K-2 in preclinical studies. The purpose of this first-in-human study is to evaluate the properties of [¹⁸F]K-40 in humans and to compare the kinetics and PET images of [¹⁸F]K-40 with those of [¹¹C]K-2. Methods: Five healthy volunteers were enrolled and underwent dynamic PET imaging using [¹⁸F]K-40 and [¹¹C]K-2. The nondisplaceable binding potential (BP_ND) with white matter as the reference was calculated by Logan graphical analysis using tissue time-activity curves (TACs), and the total distribution volume of [¹⁸F]K-40 was calculated using plasma TACs. The intraindividual correlation between BP_ND values obtained for [¹⁸F]K-40 and [¹¹C]K-2 was examined. To optimize the time window for PET scanning, BP_ND and SUV ratio were evaluated. Results: The tissue TACs of [¹⁸F]K-40 showed curves similar to those of [¹¹C]K-2. Logan graphical analysis using plasma TACs revealed reversible binding of [¹⁸F]K-40. The BP_ND obtained with [¹⁸F]K-40 and [¹¹C]K-2 significantly correlated in each corresponding region and showed very good correlation, which indicated that K-40, as observed with K-2, can provide PET images that reflect the amount of AMPARs. A good linear relationship was observed between BP_ND and the summation image of SUV ratios between 40 and 50 min after radiotracer injection. Conclusion: [¹⁸F]K-40, as with [¹¹C]K-2, has favorable binding properties as an AMPAR PET tracer. Thus, [¹⁸F]K-40 could characterize AMPAR distribution in pathophysiologic conditions of the brain and facilitate the development of novel diagnostics of neuropsychiatric disorders.