Optimizing FDG radiochemistry on digital microfluidics using ‘one-factor-at-a-time’ approach

The fluorodeoxyglucose (FDG) radiotracer is widely used as a molecular probe for Positron emission tomography (PET) imaging and plays an important role in cancer detection, diagnosis, and staging. However, the growth of PET use globally is mainly constrained by the limited availability of the radiotracer [18F]FDG due to the high infrastructure costs and the tenuous nature to synthesize these tracers for clinical use.

The long-standing method for producing radiotracers is nucleophilic fluorination (developed by Hamacher and coworkers). This requires multiple steps with different temperatures, organic solvents, and basic conditions (pH > 12) to turn [18F]fluoride into [18F]FDG. Optimization of this type of reaction that has many variables affecting their final yield is very challenging and time-consuming. Exponentially, it gets difficult by the number of variants, and it can increase the cost and experiment time to an extent that makes a full optimization impossible. Also, the synthesis time is usually long, the materials are very expensive, and researchers cannot have long exposures to the radioactive substances.

Implementing the radiosynthesis on a digital microfluidic (DMF) platform instead of the conventional kit-based production modules has shown promising results like a much shorter synthesis time and higher specificity of the final product. DMF platforms have helped in automating the reactions and reducing material consumption. But the reaction yield – i.e., the amount of radioactivity in the final product after synthesis – is still low (e.g., 20-50 %) compared to those commercial kits (about 70 %). Also, it remains very tedious and labor-intensive to study all the variables on this platform due to the limit of the number of reactions that can be performed simultaneously. 

We took a new approach to optimize the synthesis of FDG radiotracer. We had eight variables for the reaction which in the simplest manner, we had to run more than 17000 experiments to find the ideal condition. If each synthesis takes 30 min, it means we had to run the reactions for 360 days and nights non-stop.

However, we used a one-factor-at-a-time method to first find the effective variables, and then, we tested only 40 conditions around the effective variables to achieve a maximum radiochemical yield of 70 % with our DMF device. To our knowledge, this is the highest yield from on-chip synthesis reported so far.

We believe this digital microfluidic platform coupled with the one-factor-at-a-time approach can be potentially used for the optimization of other tracers, too. This will have several benefits in terms of increasing the number of patients that can be scanned per day while lowering the cost barriers for cancer-care programs paving the way to integrate PET technology in the healthcare system.