The dependence of the radiochemical yield of [¹⁸F]fluoromisonidazole (1) on different reaction parameters such as reaction time, temperature and amount of precursor was investigated for the nucleophilic substitution of tosylate by [¹⁸F]fluoride and subsequent hydrolysis of the protecting group on 1-(2'-nitro-1'-imidazolyl)-2-O-tetrahydropyranyl-3-O-toluenesulfonylpropanediol as the precursor molecule (2). Highest yields (86 % ± 6 %) were obtained using 10 mg (2) at 100°C for 10 min, whereas both at 80°C and 120°C the yields were lower (46 % ±11 % and 29 % ± 14 %, respectively). A rapid decrease of the yield was observed when the reaction time exceeded 15 min, i.e. at 100°C using 5 mg (2) the radiochemical yield decreased from 61 % ± 8 % at 15 min to 18 % ± 10 % at 60 min.

1-(2'-nitro-1'-imidazolyl)-3-fluoro-2-propanol, FMISO (1), labelled with the short-lived positron-emitter fluorine-18, was suggested as a tracer for the determination of hypoxic tissue in vivo with PET in 1984 [1]. Since that time, several synthetic approaches to FMISO have been investigated with the aim of improving radiochemical yields and simplifying the synthetic procedure in order to allow clinical applications of the tracer[2-10].

These synthetic strategies can be divided into two main groups: a nucleophilic substitution on a protected precursor with subsequent removal the protection group or epoxide ring-opening and the production of a ¹⁸F-labelled reaction intermediate epifluorohydrin with subsequent coupling to the nitroimidazole moiety under basic conditions. Both approaches have their motivation based on practical considerations. The first method tends to minimise the number of chemical transformations needed, with the aim of a high labelling efficiency and a one-pot synthesis, the second is aimed at utilising off-the-shelf chemical components and has therefore been drawing greatest attention for routine applications [6]. From a critical review of the literature the most promising direct labelling approach of FMISO seemed to be the nucleophilic substitution of the tosylate leaving group by [¹⁸F]fluoride on the tetrahydropyranyl-protected precursor

1-(2'-nitro-1'-imidazolyl)-2-O-tetrahydropyranyl-3-O-toluenesulfonylpropanediol (2), NITTP, with subsequent hydrolysis of the protecting group (Figure 1) [11, 12]. A major disadvantage that restricted a wider use of the precursor suggested by Lim et al. [11] was the time-consuming low yield synthesis of the starting material. Since recently the starting compound NITTP (2) became commercially available, this labelling approach for [¹⁸F]FMISO gained additional attractivity. However, since only very few details about the applied reaction parameters were reported so far, for routine applications of [¹⁸F]FMISO a systematic investigation of the dependence of radiochemical yield on temperature, reaction time and amount of precursor was necessary.

1-(2'-Nitro-1'-imidazolyl)-2-O-tetrahydropyranyl-3-O-toluenesulfonylpropanediol (2), NITTP, was obtained from ABX Advanced Biomedical Compounds . CH₃CN for DNA synthesis, K₂CO₃*1.5 H₂O, Suprapur, and 1M HCl were obtained from Merck. 4,7,13,16,21,24-hexaoxa-1,10-diazabicyclo [8.8.8]hexacosane (Kryptofix® 2.2.2) was obtained from Sigma-Aldrich. All reagents were used without further purification.

Labeling Procedure
[¹⁸F]Fluoride was produced at the cyclotron of the PET-Center (PETtrace, GE Medical Systems) via the ¹⁸O(p,n)¹⁸F reaction.

To 32 mg (85 µmol) Kryptofix® 2.2.2 in a 5 ml Reactivial equipped with stirring bar, Ar-supply (quality 6.0) and vacuum line were added 150 µl aqueous 0.28 M K₂CO₃ solution, 1 ml CH₃CN and 100-300 µl [¹⁸F]fluoride in H₂¹⁸O.

Excess water was removed in vacuo at 100°C and the resulting complex was dried additionally three times by azeotropic distillation with 1 ml CH₃CN each under an argon stream. A solution of 1 mg, 5 mg or 10 mg NITTP (2) in 2.5 ml CH₃CN was added and the vial was heated for 5 min, 10 min, 15 min, 30 min or 60 min to 80°C, 100°C or 120°C. From the resulting reaction mixture 100 µl were added to ice-cold 200 µl 1 M HCl. The protection group was removed by heating the resulting mixture to 100°C for 5 min. After cooling to room temperature the solution was diluted with 500 µl of HPLC eluent (H₂O/C₂H₅OH 95/5, v/v) and 100 µl were analysed by HPLC equipped with NaI-radiodetector on a Partisil 10 ODS 3 column, 500 mm*8 mm (CS Chromatographie Service), H₂O/C₂H₅OH 95/5, v/v, flow 4 ml/min, UV-detection at 320 nm. Under these conditions the k' value for [¹⁸F]FMISO was 2.3, corresponding to a retention time of 15 min. The THP-protected reaction product, 3-[¹⁸F]fluoro-1-(2'-nitro-1'-imidazolyl)-2-O-tetrahydropyranylpropanol, was not eluted from the column under the applied chromatographic conditions within 1 h. The peak corresponding to [¹⁸F]FMISO was collected and the yield was expressed as the amount of radioactivity in the [¹⁸F]FMISO fraction divided by the total injected activity as well as the area of the [¹⁸F]FMISO peak in the radioactivity channel divided by the area corresponding to an aliquot of the analysed sample injected behind the HPLC column.

RESULTS AND DISCUSSION

The product yield of [¹⁸F]FMISO was determined at reaction times between 5 min and 60 min in acetonitrile at 100°C (Figure 2). The experiments were performed with 5 mg NITTP (2) in a reaction volume of 2.5 ml CH₃CN.

Highest yields of about 60 % were obtained after 10 min and 15 min, whereas surprisingly the yield decreased with longer reaction times resulting in a yield of 38 % ± 13 % (n=5) at 30 min and 18 % ± 10 % (n=5) at 60 min, indicating a thermal instability of the reaction intermediate 3-[¹⁸F]fluoro-1-(2'-nitro-1'-imidazolyl)-2-O-tetrahydropyranylpropanol. However, no other radioactive product eluted under the applied chromatographic conditions from the HPLC column within the observed time range.

Based on these results, a reaction time of 10 min was chosen for the optimization of the reaction temperature (Table 1). The experiments were performed with 5 mg NITTP (2) in a reaction volume of 2.5 ml CH₃CN at 80°C, 100°C and 120°C. The highest yield was obtained at 100°C, at this reaction temperature 60 % ± 14 % (n=5) of the [¹⁸F]fluoride-activity were converted into [¹⁸F]FMISO. This result is in good agreement with those reported by Lim et al. [12] who found labelling yields between 43 % and 71 % under the same conditions. Both at lower and at higher reaction temperatures, i.e. 80°C and 120°C the radiochemical yield was lower, resulting in 46 % ± 11% (n=3) and 29 % ± 14 % (n=3) labelled product, respectively. The stronger decrease of the yield at 120°C can be interpreted as another hint of the assumed thermal instability of the reaction intermediate.

The dependence of the radiochemical yield on the amount of precursor was determined in a range between 1 mg and 10 mg of starting material at a reaction time of 10 min and a temperature of 100°C in CH₃CN. As shown in Table 2 radiochemical yields of 60 % ± 14 % (n=3) were obtained when 5 mg (13.2 µmol) of (2) were used in the reaction. Within the investigated range the highest yield was observed for 10 mg (26.4 µmol) of precursor. Under these conditions 86 % ± 6 % (n=3) of [¹⁸F]FMISO were produced. A dramatic decrease in the labelling yield was seen when low amounts of precursor were used for the reaction, i.e. with 1 mg of (2) the radiochemical yield was less than 1 %.

In conclusion, the labelling of [¹⁸F]FMISO by nucleophilic substitution of [¹⁸F]fluoride for tosylate and subsequent hydrolysis of the protecting group on 1-(2'-nitro-1'-imidazolyl)-2-O-tetrahydropyranyl-3-O-toluenesulfonylpropanediol (2) as suggested by Lim et al. represents an easy and reliable method that should be able to produce high amounts of [¹⁸F]FMISO since labelling yields of more than 80 % can be obtained and the reaction procedure can be performed in any automated system designed for the synthesis of FDG either using the method originally presented by Hamacher et al. [13] or the heterogenous nucleophilic substitution on a quaternary aminopyridinium resin as implemented in the GE MicroLab [8,14].

REFERENCES

1. Jerabek, P.A.; Dischino, D.D.; Kilbourn, M.R.; Welch, M.J., J. Labelled Compd. Radiopharm. 21 (1984) 1234.
2. Jerabek, P.A.; Patrick, T.B.; Kilbourn, M.R.; Dischino, D.D.; Welch, M.J., Appl. Rad. Isot. 37 (1986) 599.
3. Cherif, A.; Yang, D.J.; Tansey, W.; Kim, E.E.; Wallace, S., Pharmaceutical Research 11 (1994) 466.
4. Hwang, D.-R.; Dence, C.S.; Bonasera, T.A.; Welch, M.J., Appl. Radiat. Isot., 40 (1989) 117.
5. Grierson, J.; Mathis, C.; Shankland, E.; Link, J.; Grunbaum, Z.; Rasey, J.; Krohn, K., J. Nucl. Med., 28 (1987) 1081.
6. Grierson, J.R.; Link, J.M.; Mathis, C.A.; Rasey, J.S.; Krohn, K.A, J. Nucl. Med., 30 (1989) 343.
7. Hwang, D.-R.; Dence, C.S.; Welch, M.J.; Shelton, M.E.; Bergmann, S.R., J. Labelled Compd. Radiopharm., 29 (1989) 442.
8. Solbach, M.; Machulla, H.-J., J. Labelled Compd. Radiopharm., 37 (1995) 199.
9. McCarthy, T.J.; Dence, C.S.; Welch, M.J., Appl. Radiat. Isot., 44 (1993) 1129.
10. Tada, M.; Iwata, R.; Sugiyama, H.; Sato, K.; Kubota, K.; Kubota, R.; Takahashi, H.; Fukuda, H.; Ido, T., J. Labelled Compd. Radiopharm., 38 (1996) 771.
11. Lim, J.L.; Berridge, M.S., J. Labelled Compd. Radiopharm., 32 (1993) 541.
12. Lim, J.-L.; Berridge, M.S., Appl. Radiat. Isot., 44 (1993) 1085.
13. Hamacher, K.; Coenen, H.H., Stöcklin, G., J. Nucl. Med., 27 (1986) 235.
14. Mulholland, G.K., Mangner, T.J., Jewett, D.M., Kilbourn, M.R., Labelled Compd. Radiopharm., 26 (1989) 378.