TL;DR
Researchers have characterized a new PET tracer, [18F]GE-180, for imaging neuroinflammation in the human brain. The study found that this tracer exhibits low brain uptake and minimal influence from a common genetic variation, suggesting it might be a promising tool for future neuroinflammatory disease research, though further investigation in patient populations is needed.
What Problem Does This Paper Address?
Neuroinflammation, the activation of immune cells in the brain (microglia), is a hallmark of many neurological diseases like Alzheimer's, multiple sclerosis, and traumatic brain injury. The translocator protein (TSPO) is a key marker that becomes significantly upregulated in activated microglia, making it an attractive target for imaging neuroinflammation using Positron Emission Tomography (PET).
However, earlier TSPO PET tracers, like [11C]PK-11195, suffered from high non-specific binding and short half-lives (requiring on-site cyclotrons), limiting their clinical utility. Newer, second-generation TSPO ligands have been developed to improve signal-to-noise and specific binding. [18F]GE-180 is one such novel radiotracer, labeled with fluorine-18 (which has a longer half-life, enabling wider distribution). A significant challenge with some TSPO tracers is that their binding affinity can be influenced by a common genetic polymorphism (Ala147Thr, rs6971), leading to different binding levels in individuals classified as high-affinity binders (HABs) or mixed-affinity binders (MABs). This study aimed to:
- Characterize the kinetic behavior of [18F]GE-180 in the healthy human brain. This involves understanding how the tracer is taken up, distributed, and eliminated from brain tissue.
- Investigate if the rs6971 TSPO polymorphism affects the outcome measures obtained with [18F]GE-180.
Methods
The researchers recruited ten healthy volunteers, divided into two groups based on their TSPO genotype: five high-affinity binders (HABs) and five mixed-affinity binders (MABs). Each participant underwent a dynamic PET scan after receiving an intravenous injection of [18F]GE-180. Crucially, arterial blood sampling was performed throughout the scan to accurately measure the tracer's concentration in the blood and its metabolic breakdown (parent fraction).
The core of the analysis involved kinetic modeling of the time-activity curves (TACs) derived from different brain regions of interest (ROIs). They applied:
- One-tissue compartment (1TC) and two-tissue compartment (2TC) models: These mathematical models describe how a tracer moves between the blood and tissue over time. The 2TC model is generally more complex and can better capture tracer dynamics when it moves into and out of tissue compartments.
- Logan graphical analysis: This is a model-independent method that can estimate the total volume of distribution (VT), a measure of tracer binding, from PET data.
The primary outcome measure was the total volume of distribution (VT). Secondary measures included standardized uptake values (SUV) and their ratios (SUVR) using the cortical grey matter as a reference region (chosen because TSPO expression is typically low there in healthy individuals). They also performed time stability analysis to see if the VT estimates stabilized over longer scan durations. Statistical analysis was used to compare outcomes between the HAB and MAB groups.
Datasets
The study used data from ten healthy human volunteers. No new public datasets were introduced.
Key Insights and Ideas
- Low Brain Uptake: [18F]GE-180 shows remarkably low uptake in healthy brain tissue, with significant signal predominantly in blood vessels. This is a critical finding that influences how the tracer can be interpreted.
- Low First-Pass Extraction: The tracer has a very low extraction fraction across the blood-brain barrier (BBB), estimated at around 1%. This means only a small amount of the tracer entering the brain from the blood actually gets into the brain tissue.
- No Genotype Effect: Surprisingly, the study found no significant differences in [18F]GE-180 binding or outcome measures between individuals with high-affinity binding (HABs) and mixed-affinity binding (MABs). This contrasts with some other TSPO tracers and might be related to the low overall brain uptake.
- 2TC Model Superiority: The reversible two-tissue compartment (2TC) model provided the best fit for the PET data in most brain regions, indicating a more complex kinetic behavior than a simple 1TC model.
- Time Stability Concerns: The VT estimates did not stabilize within the 90-minute scan duration, continuously increasing. This suggests that longer scan times might be necessary for more stable VT measurements, potentially due to the accumulation of metabolites.
Key Results
- The 2TC model was deemed the most appropriate for kinetic modeling, fitting the data well.
- The initial delivery rate constant (K₁) was low across all regions (average 0.01 mL cm⁻³ min⁻¹), reflecting low extraction across the BBB.
- The median total volume of distribution (VT) was also low, ranging from 0.16 mL cm⁻³ (striatum) to 0.38 mL cm⁻³ (thalamus). These values are lower than those reported for other TSPO tracers.
- There were no significant differences in VT, SUV, or SUVR between the HAB and MAB groups across any of the ROIs.
- The plasma-to-blood ratio remained constant at about 1.69.
- The parent fraction of [18F]GE-180 in plasma was relatively high, indicating minimal rapid metabolism (65-81.7% intact at 30 min, 57.3-78.3% at 90 min).
Limitations and Future Work
The authors acknowledge several limitations:
- Low Brain Uptake: The low signal in brain tissue could reduce the signal-to-noise ratio, potentially making it harder to detect subtle changes or increasing susceptibility to bias from blood signal.
- Time Stability: The VT estimates did not reach a plateau within the 90-minute scan. This suggests that longer scan durations might be needed for more accurate and stable VT measurements, and the current estimates might be underestimates. The accumulation of radiometabolites could contribute to this ongoing increase.
- No LAB Group: The study excluded low-affinity binders (LABs) for ethical reasons, so the full spectrum of genotype effects couldn't be assessed.
- Plasma Protein Binding: The possibility of high plasma protein binding contributing to low brain delivery was raised but not directly measured.
Future work should focus on:
- Investigating [18F]GE-180 in populations with neuroinflammatory diseases. This is crucial to determine its sensitivity and specificity for detecting upregulated TSPO in conditions like Alzheimer's or multiple sclerosis.
- Exploring longer scan durations to achieve stable VT estimates.
- Further characterization of metabolites and their potential impact on brain uptake.
- Competition studies to more definitively delineate specific binding.
Summary updated