Kinetic modeling of the monoamine oxidase-B radioligand [18F]SMBT-1 in human brain with positron emission tomography

This paper describes pharmacokinetic analyses of the monoamine-oxidase-B (MAO-B) radiotracer [ 18 F]( S)-(2-methylpyrid-5-yl)-6-[(3-fluoro-2-hydroxy)propoxy]quinoline ([ 18 F]SMBT-1) for positron emission tomography (PET) brain imaging. Brain MAO-B expression is widespread, predominantly within astrocytes. Reactive astrogliosis in response to neurodegenerative disease pathology is associated with MAO-B overexpression. Fourteen elderly subjects (8 control, 5 mild cognitive impairment, 1 Alzheimer’s disease) with amyloid ([ 11 C]PiB) and tau ([ 18 F]flortaucipir) imaging assessments underwent dynamic [ 18 F]SMBT-1 PET imaging with arterial input function determination. [ 18 F]SMBT-1 showed high brain uptake and a retention pattern consistent with the known MAO-B distribution. A two-tissue compartment (2TC) model where the K 1 /k 2 ratio was fixed to a whole brain value best described [ 18 F]SMBT-1 kinetics. The 2TC total volume of distribution (V T ) was well identified and highly correlated (r 2 ∼0.8) with post-mortem MAO-B indices. Cerebellar grey matter (CGM) showed the lowest mean V T of any region and is considered the optimal pseudo-reference region. Simplified analysis methods including reference tissue models, non-compartmental models, and standard uptake value ratios (SUVR) agreed with 2TC outcomes (r 2 > 0.9) but with varying bias. We found the CGM-normalized 70–90 min SUVR to be highly correlated (r 2 = 0.93) with the 2TC distribution volume ratio (DVR) with acceptable bias (∼10%), representing a practical alternative for [ 18 F]SMBT-1 analyses.