Fluorine-18 Labelled Radioligands for PET Imaging of Cyclooxygenase-2

Molecular imaging probes enable the early and accurate detection of disease-specific biomarkers and facilitate personalized treatment of many chronic diseases, including cancer. Among current clinically used functional imaging modalities, positron emission tomography (PET) plays a significant role in cancer detection and in monitoring the response to therapeutic interventions. Several preclinical and clinical studies have demonstrated the crucial involvement of cyclooxygenase-2 (COX-2) isozyme in cancer development and progression, making COX-2 a promising cancer biomarker. A variety of COX-2-targeting PET radioligands has been developed based on anti-inflammatory drugs and selective COX-2 inhibitors. However, many of those suffer from non-specific binding and insufficient metabolic stability. This article highlights examples of COX-2-targeting PET radioligands labelled with the short-lived positron emitter 18F, including radiosynthesis and PET imaging studies published in the last decade (2012–2021).


Introduction
Cyclooxygenase (COX) enzymes play a pivotal role in the metabolism of arachidonic acid and regulate the biosynthesis of various prostanoids under normal and pathological conditions [1][2][3][4][5][6]. COX is known to exist in two main isoforms: COX-1, COX-2. Both isoforms have similar structures and show the same cyclooxygenase and peroxidase activity to produce prostaglandin H 2 (PGH 2 ), an important precursor for the biosynthesis of other prostaglandins, prostacyclin and thromboxanes. However, COX-1 and COX-2 induction, regulation, expression, and localization sites are different [7][8][9].

2012
The development of a series of 3-diarylsubstituted indole-based inhibitors was reported by Hu et al. [72]. The highly potent and selective COX-2 inhibitor 3-(4-fluorophenyl)-2-(4-methylsulfonyl-phenyl)-1H-indole (Ki(COX-2) = 20 nM; Ki(COX-1) >10 µM) was selected for the development of a novel COX-2 radioligand for PET [73]. In 2012, A major focus involves 18 F-labeled radioligands, due to their favourable half-life (t 1/2 = 109.8 min), ease of production, the availability of a variety of radiofluorination methods, and better imaging characteristics of the short-lived positron emitter 18 F. This review article primarily covers 18 F-labeled radioligands reported in the last decade for targeting COX-2. A particular emphasis is on reporting radiochemistry and PET imaging data.
In vivo PET imaging was performed in HT-29-tumor-bearing mice. During dynamic small animal PET studies, no substantial tumor accumulation of [ 18 F]-3 was observed (Figure 5). Most of the radioactivity was found to accumulate in the liver, small intestine, and kidney, and a small amount of activity (SUV ~ 1) was noticed in the brain and the blood pool. It was concluded that [ 18 F]27 is unsuitable for in vivo PET imaging of COX-2 in mice with xenotransplanted HT-29 tumors.  In vivo PET imaging was performed in HT-29-tumor-bearing mice. During dynamic small animal PET studies, no substantial tumor accumulation of [ 18 F]-3 was observed ( Figure 5). Most of the radioactivity was found to accumulate in the liver, small intestine, and kidney, and a small amount of activity (SUV~1) was noticed in the brain and the blood pool. It was concluded that [ 18 F]27 is unsuitable for in vivo PET imaging of COX-2 in mice with xenotransplanted HT-29 tumors.  Figure 4). The radiosynthesis involved the direct radiofluorination of a non-cyclized trimethylamino triflate precursor using [ 18   In vivo PET imaging was performed in HT-29-tumor-bearing mice. During dynamic small animal PET studies, no substantial tumor accumulation of [ 18 F]-3 was observed (Figure 5). Most of the radioactivity was found to accumulate in the liver, small intestine, and kidney, and a small amount of activity (SUV ~ 1) was noticed in the brain and the blood pool. It was concluded that [ 18 F]27 is unsuitable for in vivo PET imaging of COX-2 in mice with xenotransplanted HT-29 tumors.

2013
In 2013, our group reported the synthesis and in vitro evaluation of a series of trifluoromethyl-substituted pyrimidines as COX-2 inhibitors. Three fluorobenzyl-substituted pyrimidine derivatives were further advanced as 18

2013
In 2013, our group reported the synthesis and in vitro evaluation of a series of trifluoromethyl-substituted pyrimidines as COX-2 inhibitors. Three fluorobenzyl-substituted pyrimidine derivatives were further advanced as 18  The radiotracers were tested in COX-2-expressing human HCA-7 colorectal cancer cells, and in vitro specificity was examined using COX-2-negative HCT-116 cells, and by blocking studies with COX-2 inhibitors [76,77] As in the cell uptake studies results in HCA-7 cells, the in vivo studies with [ 18 F]28a ( 18 F-Pyricoxib) showed uptake in HCA-7 tumors (T/M ratio of 2.25 after 4 h p.i.) ( Figure  7). A reduced uptake of [ 18 F]28a in the tumor was noticed upon pre-treatment with the COX-2 selective inhibitor celecoxib ( Figure 8) [76,77]. However, a similar radioactivity uptake profile was observed for 18   The radiosynthesis [ 18 F]30a was accomplished using iodylaryl derivative as the labelling precursor, leading to a radiochemical yield of 5 to 10% (as determined by radio-TLC).
The radiotracers were tested in COX-2-expressing human HCA-7 colorectal cancer cells, and in vitro specificity was examined using COX-2-negative HCT-116 cells, and by blocking studies with COX-2 inhibitors [76,77] As in the cell uptake studies results in HCA-7 cells, the in vivo studies with [ 18 F]28a ( 18 F-Pyricoxib) showed uptake in HCA-7 tumors (T/M ratio of 2.25 after 4 h p.i.) (Figure 7). A reduced uptake of [ 18 F]28a in the tumor was noticed upon pre-treatment with the COX-2 selective inhibitor celecoxib ( Figure 8) [76,77]. However, a similar radioactivity uptake profile was observed for 18   In 2013, two radiolabeled COX-2 selective inhibitors, [ 11 C]celecoxib and [ 11 C] rofecoxib, were assessed as PET tracers for imaging COX-2-normal and ischemic mouse brains [66]. From a series of experiments, including in vitro autoradiography and in vivo PET assays, the authors concluded that [ 11 C]celecoxib is not a suitable COX-2 radioligand  In 2013, two radiolabeled COX-2 selective inhibitors, [ 11 C]celecoxib and [ 11 C] rofecoxib, were assessed as PET tracers for imaging COX-2-normal and ischemic mouse brains [66]. From a series of experiments, including in vitro autoradiography and in vivo PET assays, the authors concluded that [ 11 C]celecoxib is not a suitable COX-2 radioligand In 2013, two radiolabeled COX-2 selective inhibitors, [ 11 C]celecoxib and [ 11 C] rofecoxib, were assessed as PET tracers for imaging COX-2-normal and ischemic mouse brains [66]. From a series of experiments, including in vitro autoradiography and in vivo PET assays, the authors concluded that [ 11 C]celecoxib is not a suitable COX-2 radioligand for in vitro and in vivo assays, whereas [ 11 C]rofecoxib is useful for in vitro assays of COX-2.

2015
The discovery of a fluorescent probe, Celecoxib-NBD [51], in our lab led to the development of a new series of fluorine-containing cyclooxygenase-2 (COX-2) inhibitors [80]. for in vitro and in vivo assays, whereas [ 11 C]rofecoxib is useful for in vitro assays of COX-2.
The 18   Although the normalized cellular uptake of [ 18 F]31 in COX-2-positive HCA-7 cells was high (450% radioactivity per mg protein after 60 min), no noticeable blocking effect was observed by pre-treatment with known COX-2 inhibitors (celecoxib, rofecoxib). It was suggested that the uptake and retention of [ 18 F]31 in HCA-7 cells is associated with unknown non-COX-2 targets. PET imaging of radioligand [ 18 F]31 in HCA-7-tumor-bearing NIH-III mice revealed an SUVmax in HCA-7 tumors of 0.40 ±0.07 (n = 3) after 10 min p.i. (Figure 10). At the same time point, the muscle uptake was 0.29 ± 0.08 (n = 3); therefore, the tumor-to-muscle ratio was only 1.4, providing poor imaging contrast. An analysis of the time-activity curves for radioactivity uptake in the tumor suggested that radioligand [ 18 F]31 was not trapped in tumor tissue. Therefore, radioligand [ 18 F]31 is unsuitable for COX-2 imaging.  Although the normalized cellular uptake of [ 18 F]31 in COX-2-positive HCA-7 cells was high (450% radioactivity per mg protein after 60 min), no noticeable blocking effect was observed by pre-treatment with known COX-2 inhibitors (celecoxib, rofecoxib). It was suggested that the uptake and retention of [ 18 F]31 in HCA-7 cells is associated with unknown non-COX-2 targets. PET imaging of radioligand [ 18 F]31 in HCA-7-tumor-bearing NIH-III mice revealed an SUV max in HCA-7 tumors of 0.40 ±0.07 (n = 3) after 10 min p.i. (Figure 10). At the same time point, the muscle uptake was 0.29 ± 0.08 (n = 3); therefore, the tumor-to-muscle ratio was only 1.4, providing poor imaging contrast. An analysis of the time-activity curves for radioactivity uptake in the tumor suggested that radioligand [ 18 F]31 was not trapped in tumor tissue. Therefore, radioligand [ 18 F]31 is unsuitable for COX-2 imaging.

2016
A novel set of diaryl-substituted heterocycles containing a tricyclic dihydropyrrolo[3,2,1hi]indole and pyrrolo[3,2,1-hi]indole core structure were designed to improve the COX-2inhibitory activity of the previously reported compound IND [78]. From this series of compounds, two promising COX-2 selective inhibitors, DHPI (IC 50 COX-2: 0.15 mM, SI = >666) and PI (IC 50 COX-2: 0.04 mM, SI = >2500) were selected for 18 F-labelling and PET imaging studies ( Figure 11) [82]. Gassner  A series of cell uptake studies in five human cell lines which significantly differed in their COX-2 expression levels were conducted to confirm the COX-2-specific cell uptake of [ 18 F]DHPI. However, blocking experiments with the COX-2 inhibitor celecoxib did not result in a significant change in the cellular uptake of [ 18 F]DHPI. These cellular uptake study results are indicative of non-specific binding of [ 18 F]DHPI. Additional radiopharmacological experiments were conducted in female A205-tumor-bearing NMRI nu/nu mice. Consistent with cellular uptake studies, radioligand [ 18 F]DHPI uptake could also not be blocked in COX-2-positive A2058 tumors after pre-injection of celecoxib.
In another study, indomethacin, a known COX-1/2 inhibitor, was conjugated with zwitterionic phosphonium aryltrifluoroborates for radiolabeling with 18 F. The radiolabeling of these novel indomethacin conjugates was achieved by innovative 18 Figure 12). The study was focused on radiosynthesis, and PET imaging experiments were not reported.     In another study, indomethacin, a known COX-1/2 inhibitor, was conjugated with zwitterionic phosphonium aryltrifluoroborates for radiolabeling with 18 F. The radiolabeling of these novel indomethacin conjugates was achieved by innovative 18   In another study, indomethacin, a known COX-1/2 inhibitor, was conjugated with zwitterionic phosphonium aryltrifluoroborates for radiolabeling with 18 F. The radiolabeling of these novel indomethacin conjugates was achieved by innovative 18 Figure 12). The study was focused on radiosynthesis, and PET imaging experiments were not reported.

2017
Lebedev et al. utilized electrochemical radiofluorination chemistry for the radiosynthesis of 18 F-labelled COX-2 inhibitor 18 F-40 (COX-2 IC 50 = 1.5 nM) [84]. The radiolabeling occurred directly on a heteroaromatic ring. The electrochemical 18 F-labelling was performed using a radioelectrochemical synthesizer built in-house. Briefly, radiolabeling involved 18 F drying on a cartridge and pushing precursor through the cartridge into a Teflon reactor. The reaction mixture was electrolyzed at ambient temperature using 1.0 mm Pt wire electrodes.
The electrolysis and fluorination in the cell were performed for 70 min using an Autolab PGSTAT204. The slurry was subjected to purification, followed by treatment with HCl in a pre-heated reactor for 15 min.
The radiosynthesis of 18 F-40 was completed in 4 h, and the radioligand was prepared in 0.8-2% decay-corrected radiochemical yield. The molar activity was approximately 3 Ci/mmol (Figure 13).

2017
Lebedev et al. utilized electrochemical radiofluorination chemistry for the radiosynthesis of 18 F-labelled COX-2 inhibitor 18 F-40 (COX-2 IC50 = 1.5 nM) [84]. The radiolabeling occurred directly on a heteroaromatic ring. The electrochemical 18 F-labelling was performed using a radioelectrochemical synthesizer built in-house. Briefly, radiolabeling involved 18 F drying on a cartridge and pushing precursor through the cartridge into a Teflon reactor. The reaction mixture was electrolyzed at ambient temperature using 1.0 mm Pt wire electrodes. The electrolysis and fluorination in the cell were performed for 70 min using an Autolab PGSTAT204. The slurry was subjected to purification, followed by treatment with HCl in a pre-heated reactor for 15 min.
The radiosynthesis of 18 F-40 was completed in 4 h, and the radioligand was prepared in 0.8-2% decay-corrected radiochemical yield. The molar activity was approximately 3 Ci/mmol ( Figure 13). The authors observed a COX-2-dependent cell uptake of 18 F-40 in LPS-treated RAW264.7 macrophage-like cells. A reduction of radioligand uptake was detected when cells were pre-treated with the selective COX-2 inhibitor celecoxib. In vivo metabolism and PET imaging studies with radioligand 18 F-40 in healthy mice showed no retention in bones over 2 h, which is indicative of no radiodefluorination in vivo. The radioligand crossed the blood-brain barrier and showed excretion mainly through the hepatobiliary pathway. Radioligand 18 F-40 also displayed rapid blood clearance and high metabolic stability in vivo. Overall, radioligand 18     The authors observed a COX-2-dependent cell uptake of 18 F-40 in LPS-treated RAW264.7 macrophage-like cells. A reduction of radioligand uptake was detected when cells were pretreated with the selective COX-2 inhibitor celecoxib. In vivo metabolism and PET imaging studies with radioligand 18 F-40 in healthy mice showed no retention in bones over 2 h, which is indicative of no radiodefluorination in vivo. The radioligand crossed the blood-brain barrier and showed excretion mainly through the hepatobiliary pathway. Radioligand 18 F-40 also displayed rapid blood clearance and high metabolic stability in vivo. Overall, radioligand 18  thesis of 18 F-labelled COX-2 inhibitor 18 F-40 (COX-2 IC50 = 1.5 nM) [84]. The radiolabeling occurred directly on a heteroaromatic ring. The electrochemical 18 F-labelling was performed using a radioelectrochemical synthesizer built in-house. Briefly, radiolabeling involved 18 F drying on a cartridge and pushing precursor through the cartridge into a Teflon reactor. The reaction mixture was electrolyzed at ambient temperature using 1.0 mm Pt wire electrodes. The electrolysis and fluorination in the cell were performed for 70 min using an Autolab PGSTAT204. The slurry was subjected to purification, followed by treatment with HCl in a pre-heated reactor for 15 min.
The radiosynthesis of 18 F-40 was completed in 4 h, and the radioligand was prepared in 0.8-2% decay-corrected radiochemical yield. The molar activity was approximately 3 Ci/mmol ( Figure 13).  The authors observed a COX-2-dependent cell uptake of 18 F-40 in LPS-treated RAW264.7 macrophage-like cells. A reduction of radioligand uptake was detected when cells were pre-treated with the selective COX-2 inhibitor celecoxib. In vivo metabolism and PET imaging studies with radioligand 18 F-40 in healthy mice showed no retention in bones over 2 h, which is indicative of no radiodefluorination in vivo. The radioligand crossed the blood-brain barrier and showed excretion mainly through the hepatobiliary pathway. Radioligand 18 F-40 also displayed rapid blood clearance and high metabolic stability in vivo. Overall, radioligand 18     During BNCT investigations, the CCA rats treated with FBPin (20-30 mg) showed a reduction in tumor size. This therapeutic outcome of FBPin is encouraging and suggests that novel boron-containing COX-2 inhibitors should be further explored for BNCT.
Finally, with the use of boronic ester-based precursor 53, the automated radiosynthesis of [ 18 F]49 was accomplished in 46% decay-corrected radiochemical yield within 20 min ( Figure 15).
During BNCT investigations, the CCA rats treated with FBPin (20-30 mg) showed a reduction in tumor size. This therapeutic outcome of FBPin is encouraging and suggests that novel boron-containing COX-2 inhibitors should be further explored for BNCT.
Finally, with the use of boronic ester-based precursor 53, the automated radiosynthesis of [ 18 F]49 was accomplished in 46% decay-corrected radiochemical yield within 20 min ( Figure 15).   High levels of non-specific binding of [ 18 F]49 was confirmed by blocking studies with COX-2 inhibitor, as no reduced uptake of radioligand [ 18 F]49 was observed in peripheral tissues and organs. Overall, radioligand [ 18 F]49 was unsuccessful for PET imaging of COX-2 in models of neuroinflammation.
In vivo PET imaging of radioligand [ 18 F]49 in a rodent model of neuroinflammation showed uptake in several cerebral regions (cerebellum, striatum, cortex, hippocampus). However, no significant increase of [ 18 F]49 uptake in inflamed brain regions was noticed.
High levels of non-specific binding of [ 18 F]49 was confirmed by blocking studies with COX-2 inhibitor, as no reduced uptake of radioligand [ 18 F]49 was observed in peripheral tissues and organs. Overall, radioligand [ 18 F]49 was unsuccessful for PET imaging of COX-2 in models of neuroinflammation.
[ 18 F]FDF radiosynthesis was tested with several reaction conditions, and ultimately the radioligand was obtained via Lewis acid-catalyzed nucleophilic aromatic deiodo[ 18 F]fluorination reaction using tetrabutylammonium 18 F-fluoride and labelling precursor 59 (Figure 18  Radioligand [ 18 F]FDF was obtained in 7.4% decay-corrected radiochemical yield [n = 14 batches] at a molar activity of 578.8 Ci/mmol with 99.9% radiochemical purity. The overall synthesis time was 50 min. PET imaging studies with the radioligand [ 18 F]FDF in COX-1-overexpressing ovarian cancer models in mice showed COX-1-mediated uptake in the tumors compared to normal tissues. The study presented radioligand [ 18 F]FDF as a promising lead compound for further preclinical and clinical development.
PET imaging studies with radioligand [ 18 F]FMTP in normal mice demonstrated that the radioligand was able to cross the blood-brain barrier. However, the radiotracer also showed rapid brain wash-out within the first few minutes. An approximately two times higher uptake of radioligand [ 18 F]FMTP in the brain was noticed in an LPS-induced neuroinflammation model compared with PBS-treated mice.
The preliminary studies of the radioligand showed promising first results, and additional radiopharmacological studies are needed to validate and understand the COX-2specific uptake in the neuroinflammation model.
Our lab described the in situ click chemistry generation of a highly potent and selective COX-2 inhibitor (COX-2 IC50 = 90 nM, COX-1 IC50 > 100 µM) [89]. The lead compound was advanced for 18 F-labelling and PET imaging studies [90]. The Cu-mediated late-stage radiofluorination reaction with boronic acid pinacol ester-based precursor afforded [ 18 F]triacoxib in radiochemical yields of 72% (decay corrected) within 90 min, and the molar activity exceeded 90 GBq/µmol ( Figure 20). During metabolic stability analysis, ∼90% of [ 18 F]triacoxib was found intact after 60 min p.i. The radiotracer showed significant uptake in COX-2 overexpressing HCA-7 cells, but blocking studies with nonradioactive triacoxib suggest the occurrence of unidentified, nonspecific cellular uptake of [ 18 F]triacoxib in HCA-7 cells. In pre-clinical PET imaging studies (Figure 21), tumor uptake of [ 18 F]triacoxib was noticed, and the pre-treatment celecoxib (2 mg) reduced [ 18 F]triacoxib tumor uptake by ∼20% at 60 min p.i. (SUV). It was suggested that the remaining ∼80% of radiotracer uptake and retention in HCA-7 tumors is related to the nonspecific binding mechanisms, including lipophilicity, lysosomal trapping, and off-target binding not related to COX-2. Therefore, it was recommended that additional in vivo experiments should be performed to understand the nonspecific binding mechanisms of [ 18 F]triacoxib. PET imaging studies with radioligand [ 18 F]FMTP in normal mice demonstrated that the radioligand was able to cross the blood-brain barrier. However, the radiotracer also showed rapid brain wash-out within the first few minutes. An approximately two times higher uptake of radioligand [ 18 F]FMTP in the brain was noticed in an LPS-induced neuroinflammation model compared with PBS-treated mice.
The preliminary studies of the radioligand showed promising first results, and additional radiopharmacological studies are needed to validate and understand the COX-2specific uptake in the neuroinflammation model.
Our lab described the in situ click chemistry generation of a highly potent and selective COX-2 inhibitor (COX-2 IC 50 = 90 nM, COX-1 IC 50 > 100 µM) [89]. The lead compound was advanced for 18 F-labelling and PET imaging studies [90]. The Cu-mediated latestage radiofluorination reaction with boronic acid pinacol ester-based precursor afforded [ 18 F]triacoxib in radiochemical yields of 72% (decay corrected) within 90 min, and the molar activity exceeded 90 GBq/µmol ( Figure 20). During metabolic stability analysis, ∼90% of [ 18 F]triacoxib was found intact after 60 min p.i. The radiotracer showed significant uptake in COX-2 overexpressing HCA-7 cells, but blocking studies with nonradioactive triacoxib suggest the occurrence of unidentified, nonspecific cellular uptake of [ 18 F]triacoxib in HCA-7 cells. In pre-clinical PET imaging studies (Figure 21), tumor uptake of [ 18 F]triacoxib was noticed, and the pre-treatment celecoxib (2 mg) reduced [ 18 F]triacoxib tumor uptake by ∼20% at 60 min p.i. (SUV). It was suggested that the remaining ∼80% of radiotracer uptake and retention in HCA-7 tumors is related to the nonspecific binding mechanisms, including lipophilicity, lysosomal trapping, and off-target binding not related to COX-2. Therefore, it was recommended that additional in vivo experiments should be performed to understand the nonspecific binding mechanisms of [ 18 [91]. The three radiolabeled compounds were synthesized through a reaction of [ 18 F]fluoride with tosylated precursors 65a,b and [D2]65a under standard radiolabeling condition in an automated radiosynthesizer ( Figure 22). Based on their insufficient COX-2 inhibition potency, the three compounds were not forwarded for in vivo studies in tumor-xenograft-bearing mice. However, in vivo evaluation of biodistribution in healthy mice indicated that the three compounds possess similar pharmacokinetic properties. Upon 5 min p.i., the highest initial radioactivity concentration was observed in liver, adrenals, and brown as well as white adipose tissue. In contrast, dynamic PET studies indicated that [ 18 [91]. The three radiolabeled compounds were synthesized through a reaction of [ 18 F]fluoride with tosylated precursors 65a,b and [D2]65a under standard radiolabeling condition in an automated radiosynthesizer ( Figure 22). Based on their insufficient COX-2 inhibition potency, the three compounds were not forwarded for in vivo studies in tumor-xenograft-bearing mice. However, in vivo evaluation of biodistribution in healthy mice indicated that the three compounds possess similar pharmacokinetic properties. Upon 5 min p.i., the highest initial radioactivity concentration was observed in liver, adrenals, and brown as well as white adipose tissue. In contrast, dynamic PET studies indicated that [ 18 [91]. The three radiolabeled compounds were synthesized through a reaction of [ 18 F]fluoride with tosylated precursors 65a,b and [D 2 ]65a under standard radiolabeling condition in an automated radiosynthesizer ( Figure 22). Based on their insufficient COX-2 inhibition potency, the three compounds were not forwarded for in vivo studies in tumor-xenograft-bearing mice. However, in vivo evaluation of biodistribution in healthy mice indicated that the three compounds possess similar pharmacokinetic properties. Upon 5 min p.i., the highest initial radioactivity concentration was observed in liver, adrenals, and brown as well as white adipose tissue. In contrast, dynamic PET studies indicated that [ 18

2021
The encouraging outcome of a study focused on neuroinflammation imaging with the radioligand [ 11 C]MPbP (4'-[ 11 C]methoxy-5-propyl-1,10-biphenyl-2-ol) inspired the design and synthesis of a new set of honokiol analogs [92]. All four compounds were tested for their anti-inflammatory activities and compared with celecoxib. Compound    Radioligand [ 18 F]F-IV was prepared in radiochemical yield of 35% at a molar activity of 35-40 GBq/µmol. Assessment of the ex vivo biodistribution profile of radioligand [ 18 F]F-IV in LPS-treated and normal Wistar rats showed that the radioligand was able to penetrate the BBB, as high brain uptake was observed with a peak value of 2.21 ± 0.64 and 2.09 ± 0.65 (%ID/g) in the pons and medulla, respectively, at 10 min post-injection. Moreover, blocking experiments with the COX-2-selective inhibitor celecoxib showed a significant reduction of [ 18 F]F-IV uptake in almost all extracted organs and tissues of LPS rats. The most prominent reduction (20-32%) in radioactivity uptake was observed in the brain (pons and medulla), heart, lung, and kidney. Overall, the outcome of preliminary studies presents [ 18 F]F-IV as a promising candidate for COX-2-targeting PET imaging of neuroinflammation.

Conclusions
PET imaging of COX-2 has emerged as an exciting strategy for studying and understanding the role of COX-2 in inflammatory diseases and cancer.
This review summarizes the literature describing the synthesis and evaluation of COX-2-targeting PET radioligands over the last decade. The radiosynthesis of 18 F-labelled COX-2 radioligands has benefited from recent developments in 18 F radiochemistry, particularly late-stage radiofluorination. Many 18 F-labelled COX-2-targeting radioligands

2021
The encouraging outcome of a study focused on neuroinflammation imaging with the radioligand [ 11 C]MPbP (4'-[ 11 C]methoxy-5-propyl-1,10-biphenyl-2-ol) inspired the design and synthesis of a new set of honokiol analogs [92]. All four compounds were tested for their anti-inflammatory activities and compared with celecoxib. Compound

2021
The encouraging outcome of a study focused on neuroinflammation imaging with the radioligand [ 11 C]MPbP (4'-[ 11 C]methoxy-5-propyl-1,10-biphenyl-2-ol) inspired the design and synthesis of a new set of honokiol analogs [92]. All four compounds were tested for their anti-inflammatory activities and compared with celecoxib. Compound    Radioligand [ 18 F]F-IV was prepared in radiochemical yield of 35% at a molar activity of 35-40 GBq/µmol. Assessment of the ex vivo biodistribution profile of radioligand [ 18 F]F-IV in LPS-treated and normal Wistar rats showed that the radioligand was able to penetrate the BBB, as high brain uptake was observed with a peak value of 2.21 ± 0.64 and 2.09 ± 0.65 (%ID/g) in the pons and medulla, respectively, at 10 min post-injection. Moreover, blocking experiments with the COX-2-selective inhibitor celecoxib showed a significant reduction of [ 18 F]F-IV uptake in almost all extracted organs and tissues of LPS rats. The most prominent reduction (20-32%) in radioactivity uptake was observed in the brain (pons and medulla), heart, lung, and kidney. Overall, the outcome of preliminary studies presents [ 18 F]F-IV as a promising candidate for COX-2-targeting PET imaging of neuroinflammation.

Conclusions
PET imaging of COX-2 has emerged as an exciting strategy for studying and understanding the role of COX-2 in inflammatory diseases and cancer.
This review summarizes the literature describing the synthesis and evaluation of COX-2-targeting PET radioligands over the last decade. The radiosynthesis of 18 F-labelled COX-2 radioligands has benefited from recent developments in 18 F radiochemistry, particularly late-stage radiofluorination. Many 18 F-labelled COX-2-targeting radioligands Radioligand [ 18 F]F-IV was prepared in radiochemical yield of 35% at a molar activity of 35-40 GBq/µmol. Assessment of the ex vivo biodistribution profile of radioligand [ 18 F]F-IV in LPS-treated and normal Wistar rats showed that the radioligand was able to penetrate the BBB, as high brain uptake was observed with a peak value of 2.21 ± 0.64 and 2.09 ± 0.65 (%ID/g) in the pons and medulla, respectively, at 10 min post-injection. Moreover, blocking experiments with the COX-2-selective inhibitor celecoxib showed a significant reduction of [ 18 F]F-IV uptake in almost all extracted organs and tissues of LPS rats. The most prominent reduction (20-32%) in radioactivity uptake was observed in the brain (pons and medulla), heart, lung, and kidney. Overall, the outcome of preliminary studies presents [ 18 F]F-IV as a promising candidate for COX-2-targeting PET imaging of neuroinflammation.

Conclusions
PET imaging of COX-2 has emerged as an exciting strategy for studying and understanding the role of COX-2 in inflammatory diseases and cancer.
This review summarizes the literature describing the synthesis and evaluation of COX-2-targeting PET radioligands over the last decade. The radiosynthesis of 18 F-labelled COX-2 radioligands has benefited from recent developments in 18 F radiochemistry, particularly late-stage radiofluorination. Many 18 F-labelled COX-2-targeting radioligands with favourable PET imaging characteristics are based on popular selective COX-2 inhibitors like celecoxib. Instalment of 18 F was either on an aromatic ring or an aliphatic side chain. In addition, recent advancements in [ 18 F]CF 3 radiochemistry should now open new opportunities for designing and preparing COX-2-targeting radioligands at high molar activity using CF 3 -group-containing compounds like celecoxib or mavacoxib. Moreover, complementary imaging techniques like optical imaging have also helped to test novel COX-2 inhibitors and to select promising candidates for radiolabeling with 18 F.
Despite the challenges associated with high unspecific binding and metabolic stability, the preclinical data of several COX-2-targeting PET imaging agents look promising. The best candidates should be advanced for clinical testing to fully assess the potential and usefulness of COX-2 imaging with PET in the clinic.