Quasi-continuous production of highly hyperpolarized carbon-13 contrast agents every 15 seconds within an MRI system

Magnetic resonance imaging (MRI) has significantly advanced various fields, including analytical chemistry and medical diagnostics. Despite its success, MRI's sensitivity is limited due to the small fraction of spins contributing to the signal at room temperature. Clinical MRI primarily focuses on abundant substances like water and lipids, with few exceptions. Hyperpolarized contrast agents (HyCAs) have revolutionized MRI, enabling unprecedented imaging of metabolism and pH in vivo. However, the widespread adoption of HyCAs is hampered by the limitations of current production methods. Dissolution dynamic nuclear polarization (dDNP), the standard technique, is complex, expensive (polarizers cost over a million euros), and requires over an hour to produce a single dose. While advancements like the spinlab and spin-aligner polarizers have improved the production process, they still have limitations in terms of duty cycle and recovery time. Parahydrogen (pH₂)-induced polarization (PHIP) offers a significantly more cost-effective and faster alternative. PHIP methods, including hydrogenative PHIP (PASADENA, ALTADENA) and non-hydrogenative PHIP (SABRE), have demonstrated faster duty cycles, but achieving high concentrations, polarizations exceeding 10%, long polarization lifetimes (minutes), and minimal background signal in vivo remains a challenge. Continuous HP methods have emerged, focusing on continuous pH₂ supply or continuous-flow setups, such as lab-on-a-chip NMR devices. However, the integration of these methods into the MRI system for biomedical applications remains a crucial next step. A method called synthesis amid the magnet bore allows dramatically enhanced nuclear alignment (SAMBADENA) has shown promise by performing hydrogenative PHIP within the MRI system, decreasing the cost and time significantly. This study aims to achieve quasi-continuous production of ¹³C HyCAs using SAMBADENA within the MRI magnet, significantly improving the speed and efficiency of HyCA production for in vivo applications.

The literature extensively covers the challenges and advancements in hyperpolarization techniques, particularly for MRI applications. Dissolution Dynamic Nuclear Polarization (dDNP) has been the gold standard but is constrained by its cost and time requirements. This has driven the exploration of parahydrogen induced polarization (PHIP) as a faster and more affordable alternative. Different PHIP methods have been developed, such as PASADENA, ALTADENA (both hydrogenative), and SABRE (non-hydrogenative), each with varying levels of success in achieving high polarization and fast duty cycles. Numerous studies focused on optimizing PHIP parameters, including catalyst selection, reaction conditions (temperature, pressure), and spin order transfer (SOT) pulse sequences. Recent research has explored continuous-flow PHIP methods, demonstrating improvements in duty cycle but often at the expense of other critical parameters like polarization levels or concentration. The SAMBADENA method, performing PHIP within the MRI system, is highlighted as a key advancement in reducing production time and cost, by using existing MRI components. The present study builds on this foundation by aiming for quasi-continuous operation, surpassing previous PHIP duty cycles by a significant margin.

This study employed a custom-designed polysulfone reactor integrated into a 7T preclinical MRI system (Biospec 70/20, Bruker). The reactor was designed to withstand high pressure and temperature, crucial for efficient hydrogenation. A microcontroller-based system controlled fluid and gas flow, precisely synchronized with the MRI pulse program. The system uses five electronically switchable magnetic valves and five one-way in-line valves to direct the flow of gases (N₂ and parahydrogen) and the solution. The reactor integrated a water-based heating system to maintain the high temperature (80°C) needed for efficient hydrogenation. The procedure began with warming the reactor to 80°C for 5 minutes and involved filling the reactor with 1 mL of 80°C H₂O. A 10 mL syringe containing a catalyst-precursor solution was warmed to 80°C and connected to the reactor inlet. The experiment was triggered automatically, initiating the automated hyperpolarization process, repeatedly every 15 seconds. This included the steps of filling, hydrogenation, spin order transfer (using phINEPT+ for HEP and Goldman's sequence for SUC), signal detection, and emptying the reactor. Two key agents, hydroxyethyl [1-¹³C]propionate-d₃ (HEP) and [1-¹³C]succinate-d₂ (SUC), were used. The ¹³C polarization was quantified using a thermally polarized reference (acetone). The experimental parameters (temperature, pressure, catalyst concentration, precursor concentration, and hydrogenation time) were systematically optimized to maximize ¹³C polarization. A two-handed kinetics model was used to analyze the hydrogenation kinetics. A method to rapidly cool the HyCA to in vivo applicable temperatures was also tested, ensuring the agent is ready for immediate use.

This study achieved the quasi-continuous production of hyperpolarized ¹³C contrast agents (HyCAs) with a 15-second duty cycle. Using the SAMBADENA method within a 7T MRI system, mean ¹³C polarizations of (19 ± 1)% for hydroxyethyl-[1-¹³C]propionate-d₃ (HEP) and (1.7 ± 0.2)% for [1-¹³C] succinate-d₂ (SUC) were obtained repeatedly. The actual PHIP process, including hydrogenation and spin order transfer, took only 5 seconds per sample. The optimization experiments revealed that reaction temperature and parahydrogen pressure significantly influenced the hyperpolarization yield. The catalyst concentration had a less pronounced effect within the tested range. Increasing the precursor concentration improved the payload (precursor concentration multiplied by polarization) until a plateau was reached, suggesting parahydrogen became the limiting reactant. The optimal hydrogenation time was found to be 5 seconds, balancing hydrogenation yield and relaxation losses. A two-handed kinetics model was fitted to the data, indicating that (92 ± 12)% of the precursor was reacted after 5 seconds of hydrogenation. Furthermore, a method was successfully tested to cool the ejected contrast agent to approximately 32°C at the catheter tip, facilitating rapid cooling for in vivo use. This 15-second duty cycle represents a substantial improvement over existing methods (by a factor of 12 or 20 compared to previous hydrogenative PHIP methods and by two orders of magnitude compared to dDNP).

The findings demonstrate a significant advancement in the production of hyperpolarized contrast agents for MRI. The 15-second duty cycle substantially surpasses the relaxation times of ¹³C in many molecules, opening new avenues for in vivo applications. The ability to produce and potentially infuse hyperpolarized agents continuously or perform repeated injections within the agent's lifetime greatly enhances the signal-to-noise ratio through signal averaging and prolonged observation. This is particularly valuable for clinical applications of hyperpolarized MRI, where the signal is often limited. The integration of the PHIP process within the MRI system minimizes relaxation losses between polarization and administration. While purity and safety need to be ensured for in vivo use, recent advancements in rapid purification techniques offer promising solutions. The limitations of PHIP primarily relate to the solubility of parahydrogen and the kinetics of the chemical reaction incorporating the spin order into the target molecule. Further optimization could focus on using solvents with higher parahydrogen solubility to enhance the reaction rate and increase polarization.

This work successfully demonstrates quasi-continuous production of highly hyperpolarized ¹³C contrast agents with a remarkable 15-second duty cycle, using a novel setup integrated into an MRI system. The rapid and reproducible hyperpolarization achieved opens up exciting possibilities for various MRI applications including serial injections, hyperpolarized infusion, and significantly improved signal averaging. Future research will focus on extending this methodology to other metabolically active agents and exploring further improvements to the efficiency and safety of the process for in vivo use, including the incorporation of heterogeneous catalysis and advanced purification methods.

The study focused on two specific hyperpolarization agents (HEP and SUC). While the methodology is expected to be applicable to other agents, further investigation is necessary to confirm this. The optimization of the hydrogenation reaction was performed under specific conditions; variations in these conditions (e.g., different precursor concentrations or catalyst types) may impact the efficiency. The rapid cooling method was tested under specific parameters. Further investigation is needed to explore its robustness under various conditions and scalability for larger volumes. Finally, while the feasibility of continuous infusion is suggested, rigorous safety and quality assessments are required before clinical translation.