Abstract
241345
Introduction: Blood-brain barrier (BBB) permeability dysfunction is implicated in many neurodegenerative and systemic diseases. The permeability-surface area (PS) product estimated by dynamic contrast-enhanced CT or MRI is a measure of BBB permeability, but only reflects non-specific leakage of contrast material across the BBB. Quantifying the PS of PET radiotracers may open new directions to probe BBB permeability with molecular specificity, but research remains underexplored for many PET tracers (e.g., 18F-fluorodeoxyglucose, FDG) due to the complexity of existing dual-tracer methods required to measure cerebral blood flow (CBF) and BBB transport rate (K1) for PS quantification. In this study, we developed a single-tracer method using high temporal resolution (HTR) PET kinetic modeling enabled by total-body PET to quantify the molecular BBB PS of PET radiotracers and demonstrate its potential utility in characterizing molecular BBB dysregulation.
Methods: Using the first two-minutes (60×1 s, 30×2 s; 2-mm isotropic voxels) of HTR dynamic PET scans acquired on the uEXPLORER PET/CT system, we simultaneously estimated CBF and K1 with the adiabatic approximation to the tissue homogeneity (AATH) model (St Lawrence and Lee 1998) to derive PS = –CBF ln(1 – K1 / CBF) based on the Renkin-Crone equation (Renkin 1959; Crone 1963). Two human subject cohorts with IRB approval and informed consent were used to evaluate the method. First, we quantified the PS of three PET radiotracers to demonstrate the spectrum of molecular BBB PS values. Fifteen subjects were scanned with either 18F-fluciclovine (FACBC), 18F-FDG, or 11C-butanol, which were expected to have very low, moderate, and very high BBB PS, respectively (Michaud et al 2020; Hasselbalch et al 1996; Herscovitch et al 1987). Second, we quantified FDG PS to characterize molecular BBB dysregulation in patients with metabolic dysfunction-associated fatty liver disease (MAFLD) with biopsy-graded lobular inflammation and total-body dynamic FDG-PET. Differences in CBF, FDG PS, and FDG K1 were compared between patients stratified by severe versus mild lobular inflammation as well as against age-matched healthy controls by a one-way ANOVA and multiplicity-adjusted post hoc pairwise comparisons. For each subject, an image-derived input function from the ascending aorta and time-activity curves in the grey matter, white matter, and cerebellum were extracted to estimate regional CBF, K1, PS, and extraction fraction = K1 / CBF by a basis function method of the AATH model.
Results: PS, K1, and extraction fraction greatly differed while CBF appeared consistent between the three radiotracers (Figure 1A/2A). Whole-brain PS for FACBC and FDG was 0.016±0.003 and 0.132±0.010 ml/min/cm3 while that of butanol was indeterminately high due to its free diffusion across the BBB. We detected significantly lower brain FDG PS (p<0.05) in patients with severe MAFLD-associated lobular inflammation (biopsy score 2-3; N=17) compared to those with mild lobular inflammation (biopsy score 0-1; N=13) and age-matched controls (N=13; Figure 1B/2B). Significant subgroup differences (p<0.05) were also observed with BBB K1, but not with CBF or extraction fraction (p>0.05).
Conclusions: We developed a single-tracer method with HTR PET kinetic modeling to quantify the molecular BBB PS of PET radiotracers. While BBB dysfunction is conventionally associated with an increase in CT/MRI-derived PS due to vascular leakage, we characterized a decrease in FDG BBB PS in our cohort of patients with severe inflammatory MAFLD. Accordingly, the decrease in FDG BBB transport K1 appeared driven by a reduction in FDG PS rather than cerebral hypoperfusion; this may suggest a multisystemic impact of liver inflammation on the molecular dysregulation of BBB permeability, e.g., possibly via circulating cytokines. Our method opens new opportunities to non-invasively elucidate the molecular role of BBB dysregulation in the development and progression of systemic diseases such as MAFLD.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.