The new approach was investigated by Anna Celler of the University of British Columbia (Vancouver, BC) and collaborators, as part of a $35 million initiative by the Canadian government to look for alternative manufacturing techniques for Tc-99m. At present, the isotope is produced at only five nuclear reactors worldwide. The fragility of this supply chain was worryingly evident in a recent global shortage.

The use of medical cyclotrons to produce Tc-99m, if proven to be viable, offers a convenient, on-site source of isotopes for nuclear medicine departments. The cyclotrons would be used to bombard molybdenum with a proton beam, causing the transmutation of some of the nuclei into Tc-99m. However, Mo-100 is an expensive resource and the technique produces multiple contaminant isotopes that increase radiation dose to the patient and reduce the specific activity of the radiopharmaceutical, reducing the diagnostic benefit for the patient. The viability of the technique must therefore be carefully scrutinized.

This is where the theoretical modelling by Celler and her collaborators has real value. The model has enabled the researchers to predict the new manufacturing method's viability and estimate logistical parameters, such as the number of cyclotron runs needed to meet the daily demands of a typical nuclear medicine department imaging schedule. The reaction conditions needed for optimal yields, such as beam energy and target geometry, were also identified. The theoretical approach allowed yields to be predicted, while minimizing expensive experimental measurements.

The researchers used the nuclear reaction model code EMPIRE-3 to calculate the cross-section, or probability, of each of the multiple possible molybdenum-proton reactions, across an energy range of 6–30 MeV. To predict Tc-99m and contaminant yields, the researchers also developed a graphical user interface (GUI) that automated their calculation when the target isotopes, target thickness, the initial proton beam energy and intensity, and the reaction cross sections were specified, the latter generated by the EMPIRE-3 simulation.

The EMPIRE-3 simulation confirmed that the numerous molybdenum-proton reactions produced multiple contaminants, including several technetium, molybdenum, niobium and zirconium isotopes. Together with the yield calculations, the EMPIRE-3 simulation also demonstrated that only enriched molybdenum targets were viable for efficient Tc-99m production; natural molybdenum, with its composition of several isotopes, produced contaminant isotopes to such an extent that they competed with and, at times, outperformed Tc-99m yields.

The researchers also identified 16–19 MeV as the optimal energy range for Tc-99m production. In this range, relative Tc-99m yields were greatest when compared to contaminant isotopes. Shorter, multiple Mo-100 irradiation cycles per day, each 3–6 hours long, also proved to be the most efficient production schedule.

Celler and her collaborators were pleased with the study outcome. "We are very happy with these results: not only are our theoretical calculations in agreement with the existing experimental data, but also they provide us with guidance for future experiments and suggest what could be the optimal conditions for technetium production. We've also demonstrated that direct production of Tc-99m via proton irradiation of Mo-100 targets is a viable approach to the production of this important radionuclide. The yields are sufficient, so that even cyclotrons designed to produce PET radionuclides can produce sufficient quantities of Tc-99m to meet local needs."

The researchers are now using their results to calculate radiation doses to patients that will result from the cyclotron-produced technetium. "These dose calculations can then be compared with those related to reactor-produced technetium and will serve us as guidance for the selection of target enrichment," explained Celler.