📄 medicine

Imaging Monoacylglycerol Lipase A First-in-Human Positron Emission Tomography Study with [18F]MAGL-2102

This first-in-human study evaluates the novel PET radiotracer [18F]MAGL-2102 for imaging monoacylglycerol lipase, revealing that despite favorable preclinical data, its low brain uptake in humans precludes its use as a reliable tool for in vivo brain imaging, though the findings provide valuable guidance for future radioligand development.

Original authors: Bertina Jebanesan, Lucas Narciso, Nikta Zarif Yussefian, Emily Murrell, Kimberly L. Desmond, Sang Soo Cho, Jerry Warsh, Bernard Le Foll, Jeffrey H. Meyer, Junchao Tong, Jian Rong, Yinlong Li, Steven H

Published 2026-07-01

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Bertina Jebanesan, Lucas Narciso, Nikta Zarif Yussefian, Emily Murrell, Kimberly L. Desmond, Sang Soo Cho, Jerry Warsh, Bernard Le Foll, Jeffrey H. Meyer, Junchao Tong, Jian Rong, Yinlong Li, Steven H. Liang, Neil Vasdev, Isabelle Boileau

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: A "First-in-Human" Test Drive

Imagine scientists have been building a high-tech, glowing "searchlight" designed to find a specific enzyme in the human brain called Monoacylglycerol Lipase (MAGL). This enzyme is like a janitor in the brain that helps clean up chemical messengers involved in pain, mood, and memory.

For years, this searchlight (a radioactive tracer called [18F]MAGL-2102) worked perfectly in test tubes, mice, and monkeys. It lit up the brain beautifully, showing exactly where the "janitor" was working. Because it worked so well in animals, the team at the Centre for Addiction and Mental Health (CAMH) decided to take it for a test drive in real humans.

This paper is the report card from that first human test drive.

The Experiment: Six Volunteers and a Glow-in-the-Dark Scan

The researchers recruited six healthy volunteers (three men and three women). Here is what happened:

The Injection: They injected a tiny, safe amount of the glowing tracer into the volunteers' veins.
The Scan: The volunteers lay in a PET scanner (a special camera that sees radioactivity) for two hours.
The Blood Work: While the camera was taking pictures, a machine continuously drew small amounts of blood from the volunteers' arms. This was to see how the body was breaking down the tracer and how much of it was actually floating freely in the blood versus stuck to proteins.
The Safety Check: Two of the volunteers also had a full-body scan to measure how much radiation their organs (like the liver and gallbladder) received, ensuring the process was safe.

The Results: The Searchlight Fizzled Out

The team hoped the searchlight would shine brightly in the human brain, just like it did in the monkeys. Unfortunately, the results were disappointing.

The "Dim Bulb" Problem: When the tracer entered the human brain, it barely lit up. The brightness levels (called SUV) were very low, ranging from 0.66 to 0.81. To put this in perspective, a good searchlight usually needs to be around 2.0 to be useful. It was like trying to find a needle in a haystack using a candle instead of a flashlight.
The "Traffic Jam" in the Blood: Why did it fail? The researchers found that the tracer got "stuck" in the blood. About 99.3% of the tracer was clinging to blood proteins, leaving only a tiny fraction (0.67%) free to swim across the blood-brain barrier and enter the brain.
- Analogy: Imagine the blood is a highway. In monkeys, the tracer was a fast car driving freely on the road. In humans, the tracer was a car that got stuck in a massive traffic jam with other cars (proteins), so very few of them ever reached the destination (the brain).
The Metabolism: The good news is that the human body didn't break the tracer down too quickly. About 88% of it remained intact by the end of the scan. This means the tracer wasn't failing because it was being destroyed; it was failing because it couldn't get into the brain in the first place.

The Safety Report: A Clean Bill of Health

Even though the tracer didn't work for brain imaging, the safety tests were successful.

The radiation dose the volunteers received was low and well within safe limits (similar to what you might get from a few CT scans).
The tracer was mostly cleared out through the gallbladder and liver, which is a normal path for this type of molecule.
No volunteers felt sick or had any bad reactions.

The Conclusion: A Learning Experience, Not a Failure

The paper concludes that [18F]MAGL-2102 is not suitable for imaging the human brain. It simply doesn't get in.

However, the authors emphasize that this wasn't a waste of time. It was a crucial "reality check."

The Lesson: Just because a tool works in a monkey doesn't mean it will work in a human. The biology of the blood-brain barrier and how blood proteins bind to drugs can be very different between species.
The Future: This study gives the scientists a clear roadmap for what not to do next. They now know that future tracers need to be designed to avoid getting "stuck" in human blood proteins.

In short: The team built a key that fit the lock perfectly in the animal kingdom, but when they tried it on the human door, it was too big to turn. They didn't get the picture they wanted, but they learned exactly why, which helps them build a better key for the future.

Technical Summary: First-in-Human PET Study of [18F]MAGL-2102

Problem and Context
Monoacylglycerol lipase (MAGL) is a critical enzyme in the endocannabinoid system (ECS), responsible for hydrolyzing approximately 85% of the endogenous ligand 2-arachidonoylglycerol (2-AG). While MAGL inhibitors have shown promise in preclinical models for treating pain, neurodegeneration, and substance use disorders, translating these findings to humans requires reliable in vivo quantification of MAGL density. Although several PET radiotracers for MAGL have been developed, few have successfully advanced to human trials due to challenges such as poor brain penetration, lack of selectivity, or rapid metabolism. Only one prior tracer, [18F]T-401, has reached human trials, but its utility is limited by the formation of brain-penetrant radiometabolites. There is a continued need for a validated, reversible MAGL radiotracer capable of accurate in vivo quantification in humans. This study addresses the gap by evaluating [18F]MAGL-2102, a novel reversible radiotracer previously characterized in rodents and non-human primates (NHP), in a first-in-human setting.

Methodology
The study was a first-in-human evaluation involving six healthy participants (3 males, 3 females; mean age 27.5 ± 10.7 years) and two additional participants for dosimetry analysis.

Radiosynthesis: [18F]MAGL-2102 was synthesized using an automated module at the Centre for Addiction and Mental Health (CAMH). The process involved nucleophilic substitution of a boronic ester precursor with [18F]fluoride, followed by HPLC purification. The final product met all quality control criteria for human administration, including radiochemical purity ≥90%, molar activity of 139 ± 36 GBq/µmol, and endotoxin levels <17.5 EU/mL.
Imaging Protocol: Participants underwent dynamic PET brain scans (120 minutes) on a GE Discovery MI scanner following an intravenous bolus injection of approximately 180 MBq of [18F]MAGL-2102. Arterial blood sampling was performed continuously for the first 22 minutes and manually at specific intervals to generate arterial input functions (AIFs).
Metabolite and Plasma Analysis: Plasma samples were analyzed via HPLC to determine the parent fraction and radiometabolites. The free fraction ( $f_p$ ) of the tracer in plasma was also measured.
Kinetic Modeling: Brain PET data were processed using PMOD software. Regional time-activity curves (TACs) were generated for 14 volumes of interest (VOIs). Kinetic modeling included Logan graphical analysis (using a metabolite-corrected AIF), one- and two-tissue compartment models (1TCM, 2TCM), and a simplified reference tissue model (SRTM2) using the brainstem as a reference region.
Dosimetry: Two participants underwent whole-body dynamic PET/CT scans to estimate radiation absorbed doses and effective doses using OLINDA/EXM 2.3 software.

Key Results

Safety and Synthesis: The radiosynthesis was robust, yielding high radiochemical purity. No adverse events were reported following administration.
Pharmacokinetics: The tracer showed minimal metabolism in human plasma, with a parent fraction of approximately 88 ± 4.4% at the end of the scan. However, the free fraction of the tracer in plasma was very low (0.67 ± 0.03%).
Brain Uptake: Brain uptake was notably low. Standardized Uptake Values (SUV) in cortical regions ranged from 0.66 to 0.81 g/mL, with the highest uptake in the occipital lobe (0.81 ± 0.12 g/mL) and the lowest in the brainstem (0.30 ± 0.07 g/mL).
Kinetic Parameters: The total distribution volume ( $V_T$ ) was low across all regions, with a whole-brain average of 0.13 ± 0.02 mL/cm³. The highest $V_T$ was observed in the occipital lobe (0.16 ± 0.03 mL/cm³) and the lowest in the brainstem (0.05 ± 0.01 mL/cm³). When normalized to the free fraction ( $V_T/f_p$ ), values ranged from 9 to 28 mL/cm³.
Modeling Consistency: Logan graphical analysis showed linearization after 45 minutes, consistent with reversible kinetics. SRTM2 and Logan reference tissue models produced comparable Distribution Volume Ratio (DVR) values (average ~2.4), confirming the reliability of the kinetic estimates despite low signal.
Dosimetry: The gallbladder received the highest absorbed dose (45.6 µGy/MBq in males, 81.3 µGy/MBq in females), indicating hepatobiliary clearance. The estimated effective doses were 2.1 mSv for the male and 3.1 mSv for the female participant, which are within the typical range for 18F-labeled radiopharmaceuticals.

Significance and Conclusions
The authors conclude that despite promising preclinical data in rodents and NHPs, [18F]MAGL-2102 demonstrates low brain uptake and low $V_T$ values in humans, rendering it unlikely to serve as a reliable tool for imaging MAGL in vivo in the human brain. The study highlights a significant species discrepancy: while the tracer reached equilibrium in NHPs, it plateaued almost immediately in humans.

The paper attributes the low human brain uptake primarily to the extremely low free fraction of the tracer in plasma (0.67%), likely due to high plasma protein binding associated with the tracer's lipophilicity (Log D = 3.68). This suggests that species-specific differences in plasma protein binding and metabolism may critically impact the translation of MAGL radiotracers from preclinical models to humans.

The significance of this work lies not in the success of the tracer, but in the rigorous characterization of its failure. The authors posit that these findings provide essential guidance for future MAGL radioligand discovery. Specifically, they suggest that incorporating plasma free fraction and protein binding analysis in both NHP and human blood samples prior to in vivo imaging could improve the early evaluation of tracer suitability. This study establishes a pharmacokinetic and metabolic baseline for [18F]MAGL-2102 in humans, informing the development of next-generation radiotracers for the endocannabinoid system.