Synthesis and evaluation of 4-[18F]fluorothalidomide for the in vivo studies of angiogenesis

In this study, we prepared 2-(2,6-dioxopiperidin-3-yl)-4-[18F]fluoroisoindole-1,3-dione (4-[18F]fluorothalidomide; [18F]1) for the in vivo studies of angiogenesis. Radiochemical synthesis of [18F]1 was carried out by labeling 4-trimethylammoniumthalidomide trifluoromethanesulfonate with nBu4N[18F]F in dimethyl sulfoxide (DMSO), followed by reverse-phase HPLC purification. Decay-corrected radiochemical yield of [18F]1 was 50–60%, with an effective specific activity of 42–120 GBq/μmol (end of synthesis). Incubation of the radioligand with human umbilical vein endothelial cells (HUVEC-C; American Type Culture Collection) showed a time-dependent increase in the uptake of the radioligand, and the uptake was inhibited by 8–11% in the presence of 10 μM thalidomide, indicating nonspecific binding of the radioligand. Positron emission tomography (PET) images of mice implanted with tumors in their right flanks revealed a marked accumulation of radioactivity in the livers, kidneys and bladders of the mice, and brain uptake appeared at approximately 40 min after injection. However, no radioactivity uptake was detected in the implanted tumor. Thin-layer chromatography (TLC), HPLC and LC-MS analyses of mouse liver microsomal metabolites of [18F]1 and 1 with or without nicotinamide adenine dinucleotide phosphate (NADPH) clearly revealed that the radioligand did not go through metabolic activation but underwent nonenzymatic hydrolysis at physiological pH. Therefore, these results would appear to indicate that [18F]1 may not be suitable for the in vivo studies of angiogenesis at least in mice, although it was reported that thalidomide and/or its hydrolysis products may be responsible for its activity in humans.