Building a database for brain 18kDa translocator protein (TSPO) imaged using [11C]PBR28 in healthy subjects

• Published in: The Journal of nuclear medicine (1978) Vol. 58; p. 339
• Main Authors: Paul, Soumen, Liow, Jeih-San, Gallagher, Evan, Zoghbi, Sami, Gunn, Roger, Fregonara, Paolo Zanotti, Morse, Cheryl, Pike, Victor, Innis, Robert, Fujita, Masahiro
• Format: Journal Article
• Language: English
• Published: New York Society of Nuclear Medicine 01.05.2017
• Subjects: Affinity; Age; Aging; Binding; Body size; Brain diseases; Brain stem; Correlation; Correlation analysis; Gender aspects; Gender differences; Hammers; Image resolution; Image segmentation; Magnetic resonance imaging; Medical imaging; Polymorphism; Proteins; Target detection; Target recognition; Veins & arteries
• ISSN: 0161-5505; 1535-5667

Objectives: Neuroinflammation plays an important role in a number of brain disorders. Translocator protein (TSPO) has been widely used as an imaging target to detect inflammation, and [11C]-PBR28 is currently one of the mostly widely used ligands to image this target. This study sought to build a database of TSPO imaged in healthy subjects with [11C]-PBR28 that could be used to: 1) compare patient data; 2) estimate specific binding; 3) study the effects in healthy subjects of potential contributors to neuroinflammation (i.e., age, sex, and obesity); and 4) determine how many arterial plasma samples are required to accurately measure distribution volume (VT). Methods: [11C]-PBR28 scans conducted in 48 healthy subjects (23 high-affinity binders (HABs) and 25 mixed-affinity binders (MABs), 20F/28M, age: 40.9±16.9 years) were retrospectively analyzed. Seventy-eight volumes of interest (VOIs) were defined based on high resolution magnetic resonance imaging (MRI) by PNEURO/PMOD, which uses segmentation, spatial normalization, and the Hammers' N30R83 maximum probability atlas. In each scan, whole brain VT was measured via three kinetic models: unconstrained two compartment model, Logan plot, and Ichise multilinear analysis (MA1) using a radiometabolite-corrected arterial input function composed of 24 samples. Non-displaceable VT (VND) was estimated via the polymorphism plot (1) using differences in VT between HABs and MABs. Correlations between VT and age or body mass index (BMI) were examined by including data from both HABs and MABs, controlling for affinity type (i.e., partial correlation). Gender differences in VT were investigated by controlling for affinity type and age. VT was also calculated from six, 12, and 18 arterial samples. Results: Among the three kinetic models, Logan plot provided the most precise VT values. Logan plot did not give non-physiologic VT in any region and underestimated VT only 5.5% relative to the two-compartment model. The two-compartment model did not identify VT well in 10% of datasets with SE >10%. Ichise MA1 yielded similar, but slightly poorer, results to the Logan plot; it underestimated VT relative to the two-compartment model by 7.5% and gave non-physiologic VT in four regions. Given these results, Logan plot was used for subsequent analyses. VT values by Logan plot were 3.7±0.8 and 3.0±0.8 for HABs and MABs, respectively, across 78 VOIs. VT values were fairly uniform throughout brain regions; the highest binding was in the brain stem, with moderate binding observed in most cerebral cortices. The polymorphism plot estimated VND as 1.4 (95% CI 0.6-1.8). Specific-to-nondisplaceable ratio (BPND) was 1.6±0.6 and 1.1±0.6 in HABs and MABs, respectively. VT increased significantly with age in all except four regions. In addition, VT was well estimated with a smaller number of samples; specifically, VT from six and 12 samples was highly significantly correlated with VT calculated from 24 samples(r = 0.9 for both six and 12 samples) , with an overestimation of 9 and 1% for six and 12 samples, respectively. Conclusion: Logan plot provided precise binding measurements for all 78 areas of Hammers' maximum probability atlas and provided data for normal TSPO distribution. HABs showed 1.4 fold greater levels of specific binding than MABs. Older subjects showed significantly greater levels of VT consistent with the presence of aging-related inflammation. Despite some overestimation, six to 12 arterial samples sufficed to study group differences in TSPO levels.