UCLA

Positron emission tomography (PET) is a powerful medical diagnostics and research tool that uses radiolabeled molecules (tracers) to image biological processes in vivo. By administering only nanomolar quantities of the tracers, PET scans enable non-invasive assessment of normal biological processes in cells and their failure in disease to aid in medical diagnostics, staging of disease severity and monitoring treatment response. Short-lived radioisotopes used in the synthesis of diagnostic PET tracers necessitate that the tracer production is carried out shortly before imaging. Each production is a multistep process involving acquisition of the radioisotope, radiochemical synthesis and quality control. Radiochemical synthesis is further broken down into radiochemical reaction to link the radioisotope with a ligand, purification and formulation to obtain pure injection-ready product.Despite impressive sensitivity and accuracy of PET in medical diagnostics, access to the wide variety of short-lived radioactive tracers is hindered due to a high cost of their preparation. The overall preparation cost covers: (i) radiosynthesis units and various consumables, (ii) large shielded fume hoods (hot-cells), (iii) reagents and radioisotope, (iv) labor and safety. A typical centralized production of PET tracers in large radiochemistry facilities alleviates the high cost per production by splitting of the large batches and shipping those to multiple users, which is only possible for tracers in high demand such as [18F]fluorodeoxyglucose ([18F]FDG). Newly introduced dose-on-demand concept proposes convenient production of any tracer of interest directly at the imaging location, leading to a decentralized approach. While setting up dedicated conventional radiosynthesis modules for this purpose is unrealistic due to the high cost and large footprint, microfluidic approaches provide a path for a miniaturized, economical tracer production. Our lab among others has been extensively working on the radiosynthesis miniaturization using droplet microfluidic methods. Microfluidics offers potential of cost reduction in almost all aspects of radiosynthesis. (i) The synthesizer cost can be reduced as a result of reduced system size, complexity and consumables cost, (ii) the compact units can be self-shielded not requiring use of hot-cells, (iii) the reagent and radioisotope consumption per synthesis is orders of magnitude less, (iv) the faster synthesis times and small-scale production levels reduce labor burden and improve safety. Additionally, in fluorine-18 radiochemistry, the conventional synthesizers must use high starting radioactivity to ensure good molar activity (radioactivity per moles of the substance) of the final product. Previously, our group has developed automated droplet reactors based on electrowetting on dielectric (EWOD) technology. However, those devices had limitations due to high cost and complexity of fabrication. More recently, we introduced low-cost disposable silicon chips which serve as a reactor within an ultra-compact automated radiosynthesizer unit, comparable in size to a 12 oz. coffee cup. The major part of this dissertation was adapting various conventional fluorine-18-labeling synthesis methods to a droplet format to demonstrate the versatility of this radiosynthesis approach. For a number of tracers, we describe synthesis miniaturization process and demonstrate that the droplet syntheses exhibit higher yields and improved synthesis times in comparison to the conventional methods. All syntheses feature orders-of-magnitude reduction in reagent consumption and are able to achieve high molar activity even when lower starting amounts of the radioisotope are used. As a result, low cost per production can be achieved and such methods are readily useable for research in particular that involves small animal PET imaging. Importantly, we further demonstrate that these syntheses are scalable, and that production of a few human doses is feasible. For validation, we also perform full clinical quality control on these tracer batches. These results suggest that in the future it would be possible to introduce this technology in a clinical setting as well for an easy access to a wide variety of PET tracers. From the radiochemistry standpoint, microfluidic fluorine-18 radiolabeling has been shown to work for different type of ligands. Apart from small molecule radiolabeling, peptide labeling routes via droplet radiochemistry are also shown. In this work we feature isotopic exchange (IEX) fluorination that has a benefit of simplified purification but is challenging on conventional scale due to inherently lower molar activity in this type of reactions. We demonstrate that the droplet approach with its small volumes allows one to perform IEX synthesis of trifluoroborate-based peptides and prosthetic groups with high yields and molar activities. This dissertation presents reliable methods for fluorine-18-labeled radiopharmaceutical production on demand, in a time- and cost-efficient manner for diverse PET tracers. While with the current progress, these methods can be readily applied for research purposes including pre-clinical PET imaging, and future work will focus on improvements and optimization of the droplet microfluidic technologies to advance these methods to the clinical PET.