Clinical Evaluation of PET Radiotracers

National Institute Of Mental Health

Investigators

Linked publications, trials & patents

Abstract

Studies were conducted under the following NationalClinicalTrials protocols: NCT02743377, NCT03912428, NCT03324646, NCT04396873, NCT03958630, and NCT04582916. 1) Evaluation of novel PET radioligands for imaging cyclooxygenase (COX)-1 and COX-2 For several years, we have focused on developing much-needed, effective, sensitive, and reliable radioligands to image neuroinflammation. For instance, in 2017, we developed radioligandsnow widely usedfor the 18 kDa translocator protein (TSPO). Our recent efforts have focused on the cyclooxygenase (COX) system, which comprises two isoforms, COX-1 and COX-2. Both are important targets for neuroinflammatory biomarkers in neuropsychiatric disorders. Recognizing that in vivo imaging of COX-1 and COX-2 could be a powerful tool for assessing neuroinflammatory response, we recently developed two PET radioligands: 11CPS13 for COX-1 and 11CMC1 for COX-2. In human and nonhuman primate whole blood assays, PS13 was highly potent and selective for COX-1 compared to COX-2, and MC1 was highly potent and selective for COX-2 compared to COX-1. As reported in past annual reports, we previously found that 11CPS13 had excellent in vivo selectivity in animal studies. With regard to 11CMC1, we previously found that it was the first PET radioligand to successfully image and quantify COX-2 upregulation in living nonhuman primate brain after lipopolysaccharide (LPS)-induced neuroinflammation. In a subsequent study, higher 11CMC1 uptake was observed in the symptomatic joints of patients with rheumatoid arthritis compared to healthy controls. In nonhuman primate brain, specific binding of 11CMC1 was markedly increased after repeated LPS injection, while that of 11CPS13 was unchanged. Building on this work, we investigated whether 11CPS13 could be used to quantify constitutive levels of COX-1 in healthy human brain. Brain test-retest scans with concurrent arterial blood samples were obtained in 10 healthy participants of both sexes. In the brain, 11CPS13 showed highest uptake in the hippocampus and occipital cortex. There was no radiometabolite accumulation, and shorter scans were able to reliably measure distribution volume. A significant correlation was observed between distribution volume and COX-1 transcript levels, indicating that 11CPS13 binding reflects actual COX-1 density in the human brain. These results from the first-in-human evaluation of the ability of 11CPS13 to image COX-1 in the brain justifies extending the study to disease populations with neuroinflammation. For instance, because 11CPS13 selectively binds to COX-1 over COX-2, it can be used to measure the in vivo potency of NSAIDs to inhibit radioligand binding to COX-1 in organs of the body. Taken together, this work suggests that, in both nonhuman primates and humans, 11CPS13 and 11CMC1 can be used to measure neuroinflammation. As the first radioligands to act directly on COX-1 and COX-2, 11CPS13 and 11CMC1 can be used to measure COX-1 in healthy conditions and COX-2 in inflammatory disorders. They can also be used to assess drug delivery and in vivo selectivity in therapeutic trials of NSAIDs. Together, they provide a powerful new research tool for the study of mental disorders associated with neuroinflammationincluding dementias and MDD. 2) Development of 18FPF-06445974 to quantify phosphodiesterase 4B in living human brain. Phosphodiesterase-4 (PDE4) metabolizes and thereby terminates the actions of the second messenger cyclic adenosine monophosphate (cAMP). Rolipram is a reversible PDE4 inhibitor, and 11C(R)-rolipram binding is positively correlated with cAMP signaling. As described in previous annual reports, our laboratory found that 11C(R)-rolipram binding was globally decreased in unmedicated patients with major depressive disorder (MDD) experiencing a major depressive episode compared to healthy volunteers. As a follow-up, we found that, in individuals with MDD, two months of treatment with an SSRI antidepressant increased (normalized) 11C(R)-rolipram binding compared to pretreatment values. 11C(R)-rolipram binds to all four subtypes of PDE4: 4A, 4B, 4C, and 4D. While we previously found that the sum of all PDE4 subtypes was decreased during a major depressive episode, some subtypes could be decreased and others increased even though the sum total of all subtypes was decreased. We thus sought to establish whether PDE4B, specifically, is decreased during a major depressive episode in unmedicated MDD participants. Such a result would provide a rationale to test a PDE4B inhibitor in MDD patients in the future. In collaboration with Pfizer and the Karolinska Institutet, we developed the novel radioligand 18FPF-06445974, which is preferential for PDE4B in brain, and investigated its characteristics in rats, nonhuman primates and, for the first time, humans. A concomitant goal was to determine whether 18FPF-06445974 could accurately quantify PDE4 in human brain. Two human participants (one male, one female) underwent two whole body PET scans for radiation exposure dosimetry and three brain PET scans for specific PDE4B quantification. In one male nonhuman primate, four PET whole body scans were performed on different days. The first scan was acquired under baseline conditions, and rolipram (1.0 mg/kg), a PDE4 inhibitor, was administered IV in the second scan. The same baseline and blocked paradigms were repeated with PF-06445974 (0.1 mg/kg) as the blocking drug. Finally, brain PET scans were conducted in four rats. 18FPF-06445974 readily entered human brain and showed widespread distribution, with lower binding in cerebellum. Distribution volume was highest in thalamus and striatum, broadly matching human brain RNA expression. Whole-body imaging in the two humans and one nonhuman primate showed early distribution in the blood pool, accumulation in target organs, and excretion via urinary tract. Both drugs substantially blocked brain uptake in nonhuman primate blocking scans; at 60 minutes, rolipram blocked 78% of uptake, and PF-06445974 blocked 92%. However, for the rolipram-blocked scan, radioactivity in brain increased after 20 minutes and returned to baseline by the end of the scan. Rats showed a similar pattern. To assess whether this was due to radiometabolite in brain or rapid rolipram clearance, a second rolipram dose was injected as a displacing agent at 120 minutes in a rat that had undergone pre-injection of 1 mg/kg iv. This repeat rolipram dose replenished rolipram concentrations in brain and restored its ability to block radioligand binding, proving that no radiometabolite entered rat or nonhuman primate brain. Taken together, the findings demonstrated that 18FPF-06445974 successfully imaged and quantified PDE4B in human brain without radiometabolite contamination, establishing it as a valuable tool for future neuropsychiatric studies. Work is ongoing to investigate whether PDE4B levels are elevated in the brains of individuals with MDD experiencing a major depressive episode compared to healthy volunteers, and whether these levels correlate with clinical rating scales, an index of disease severity.