Feasibility of F-18 radiolabeled brain-penetrable bi-specific antibody radioligands for in vivo PET imaging of tauopathy†

Abstract

PET imaging offers promise for earlier detection and prognostication of Alzheimer's disease. Recently, antibody-based constructs that penetrate the CNS via the transferrin receptor (TfR) have improved tau-selectivity, something that currently limits small molecule tau PET radiotracers. However, it remains unclear if the slow pharmacokinetics of these constructs (MW >50 kDa) limit target binding detection within the time window available for an F-18 based radiotracer. We synthesized three radio-probes by conjugating [¹⁸F]SFB with individual bi-specific antibody constructs: 1) a full-size IgG tau antibody conjugated with a TfR fragment (TAUb), 2) a tau-scFv bispecific antibody (TAUs), and 3) an Aβ-scFv bispecific antibody (Aβs). We scanned a series of sex- and age-matched wild-type (WT) and transgenic mice with tauopathy (PS19). Each paired study consisted of three sets of PET/CT scans: an initial low dose dynamic scan followed by two static scans at 8 h and 12 h after injection of a high dose of the same probe. For TAUs probes, the whole brain uptake was higher in PS19 mice (0.0684 ± 0.0273% ID cc⁻¹, n = 5) compared to WT (0.0513 ± 0.0197% ID cc⁻¹, n = 4) though the difference did not reach statistical significance (p = 0.56). Regional quantification analysis provides supporting evidence that TAUs displayed higher specific binding over Aβs in brain regions of PS19 mice. There was net accumulation of all three probes between 8 h and 12 h, suggesting that F-18 radiolabeled bi-specific antibody constructs may not adequately quantitate deposition of tau aggregates within the available time window for F-18, limited by slow pharmacokinetics and lack of a suitable reference region.