PET Quantification in Healthy Humans of Cyclooxygenase-2, a Potential Biomarker of Neuroinflammation

• DOI: 10.2967/jnumed.124.268525
• Publication Date: 2025-01-23T19:45:15Z
• Authors: Xuefeng Yan; Martin Noergaard; Cheryl L. Morse; Jeih-San Liow; Jinsoo Hong; Douglas Greve; Sanjay Telu; Min-Jeong Kim; Jose A. Montero S...; Anthony Galassi; Ningping Feng; Sarah K. Williams...; Ted B. Usdin; Shawn Wu; Andrea Zhang; Lester S. Manly; Madeline Jenkins; Maia Van Buskirk; Adrian Lee; Sami S. Zoghbi; Victor W. Pike; Paolo Zanotti-Fre...; Robert B. Innis
• Keywords: PET; COX-2; human brain; neuroinflammation

Cyclooxygenase-2 (COX-2) is present in a healthy brain at low densities but can be markedly upregulated by excitatory input and inflammogens. This study evaluated the sensitivity of PET radioligand [¹¹C]-6-methoxy-2-(4-(methylsulfonyl)phenyl)-<i>N</i>-(thiophen-2-ylmethyl)pyrimidin-4-amine ([¹¹C]MC1) to detect COX-2 density human brain. <b>Methods:</b> The specificity [¹¹C]MC1 was confirmed using lipopolysaccharide-injected rats transgenic mice expressing <i>COX-2</i> gene, with 120-min baseline blocked scans COX-1 selective agents. Twenty-seven participants were injected [¹¹C]MC1. Ten these received 2 scans: followed blockade celecoxib (600 mg orally), preferential inhibitor. Seventeen underwent test–retest imaging. All included concurrent arterial sampling. tissue-to-plasma ratio equilibrium (i.e., total distribution volume) determined 2-tissue compartment model (2TCM). <b>Results:</b> In humanized mice, 70%–90% uptake nonradioactive MC1 (a inhibitor) not PS13 inhibitor), thereby confirming specific binding COX-2. Radioactivity peaked concentration about 4.0 SUV, indicating good passage through blood–brain barrier. Values for volume achieved stability after 80 min, no radiometabolite contamination. Celecoxib reduced neocortical areas 25% had little or effect subcortical regions cerebellum, which correlated messenger RNA expression levels. Binding site occupancy virtually complete, as Lassen plots. Test–retest reliability moderate (intraclass correlation coefficient, 0.65) relatively large variability (absolute retest variability, 20%). Reference tissue methods yielded results comparable those 2TCM up 75% intersubject (coefficient variation) half. Thus, compared 2TCM, requires blood, reference method expected significantly reduce sample sizes required statistically significant differences between groups. <b>Conclusion:</b> has adequate measure brain, suggesting it also quantify elevations disorders associated neuroinflammation.