The increasing use of quantitative imaging in the understanding of the pathophysiological processes, monitoring disease progression, and assessment of responses to treatment has become evident in modern healthcare in the past decade. Using short-lived radionuclides, positron emission tomography (PET) harnesses the ability to visualize and quantify specific biochemical processes in vivo non-invasively. However, despite the increasing importance of PET, cost-restrictive and complex production is undermining due to the use of highly specialized facilities needed for conventional macroscale radiosynthesis systems. To reduce the cost associated with PET, centralized production of large batches of radiopharmaceuticals has been leveraged to achieve economies of scale. While cost-effective, this production pathway severely limits the diversity of PET radiopharmaceuticals that are available for routine use. In recent years, microfluidic technologies have hinted at the possibility replace macroscale vial-based reactors and to restructure the current centralized production of PET radiopharmaceuticals. Microvolume syntheses, allow the reduction of reagents utilized, reduce the overall synthesis time, and in some cases have led to higher yield. While many initial advances were achieved with micro-flow devices, and micro-batch devices, the complexity of operation, poor robustness, and high cost of prototyping have led to limited use. Instead, our group has recently pioneered the use microfluidic reactors that allow the use of sessile droplets for microvolume radiosyntheses. With these devices our lab has improved radiochemical yield, lowered reagent use, allowed high molar activity to be achieved with low reaction activities, shortened synthesis time, all in a compact and low-cost apparatus. These initial studies highlighted the possibility of clinical scale production. Despite these advances, there are some aspects that require further improvement before widespread use. The current reliance on conventional chromatography systems with a large footprint, undermine the advances in miniaturization and cost-reduction of the droplet microfluidic system. Second, though clinical-scale production has been achieved, it Is not as reliable or easy to use as other aspects of the droplet system. Third, the adaption of tracer synthesis protocols to this platform requires tedious optimization since reaction conditions at the microscale are usually different from conventional vial-based reactors. Several advances are necessary to allow the advances seen with the sessile microvolume droplet reactors pioneered by our group to achieve the cost effective on demand syntheses of PET radiopharmaceuticals.
We describe here several advances toward addressing these shortcomings. One development is the creation of new high-throughput methodologies that allow the synthesis of PET radiopharmaceuticals in a massively parallel fashion, with dozens of times greater throughput than possible with conventional technology, allowing the fine tuning of reaction parameters to increase yields, reduce reagent consumption, and shorten reaction times. In conjunction, we developed high-throughput methodologies for the analysis of radiopharmaceuticals via radio-TLC, and the development of approaches (PRISMA) to achieve high chromatographic resolution on par with conventional HPLC systems.
Once an optimized droplet synthesis protocol is achieved, it can be immediately used for routine production in a droplet-based radiosynthesizer. We explore novel, more robust methods of radionuclide concentration that have allowed scaling to clinically-relevant batches of a diverse range of radiopharmaceuticals. We have also developed miniaturized methods for purification and formulation based on high-resolution radio-TLC that can potentially replace HPLC-based purification protocols. This novel method enabled compact, fast, and high-efficiency purification of PET radiopharmaceuticals, and, furthermore, eliminated the need for a separate formulation step since the purified radiopharmaceutical is provided in a biocompatible buffer solution, enabling further simplification of the system and reduction of overall synthesis time.
The improved radio-TLC methods were also applied to the assessment of the metabolism of radiopharmaceuticals in tissue such as blood. Compared to HPLC, SPE, or earlier TLC analysis approaches, we achieved high sensitivity by using commercially available channeled TLC plates with concentrating zones to permit the deposition of large sample volumes without adversely affecting resolution. In fact, we showed for the first time that TLC-based analysis has sufficient sensitivity for analysis of clinical plasma samples to profile the metabolism of [18F]FEPPA as a function of time.
In addition to the microreactor systems, we also developed novel radiolabelling techniques. Experimental approaches with electrochemistry are also described for the fluorination of electron rich moieties that are not able to be directly labelled through conventional radiochemistry strategies.
Finally, we describe the development of synthesis protocols for the routine production of a potentiometric PET radiopharmaceutical, [18F]FBnTP, to study mitochondrial membrane dysfunction in the context of lung cancer metabolism. Initial studies reveal that tumors feature metabolic heterogeneity not assessable through conventionally accepted clinical profiling. Further studies suggested that tumors with high oxidative phosphorylation (OXPHOS), feature distinct compartments of mitochondrial networks. This data suggests in lung cancers, mitochondrial networks are compartmentalized into distinct subpopulations that govern bioenergetic capacity. These studies elucidate a potential treatment pathway for tumors, that harness mitochondrial complex inhibitors. Our group is currently surveying microscale radiosynthesis methods for the production of [18F]FBnTP to leverage the high interest of the community in this radiopharmaceutical. Not described in this dissertation, we conducted further studies to develop novel potentiometric analogues of [18F]FBnTP that could measure mitochondrial dysfunction in the brain.
