Summary:

Liposomes are an established delivery system for therapeutics, but liposome production methods have suffered from issues relating to quality and scalability that must be overcome before liposomes can be widely employed in clinical applications. Microfluidics is a relatively new method of manufacturing liposomes with the potential to overcome the quality and scalability issues inherent in previous methods. The Perrie lab, from Aston University in the UK, took it upon themselves to investigate how the parameters of a microfluidic process can be adjusted to reliably control liposome size and how liposome characteristics can be maintained during high throughput manufacturing. Using a NanoAssemblr™ Benchtop, they validated that microfluidics is a viable high-throughput method for the scalable and highly reproducible manufacture of size-controlled liposomes and developed a statistical model to show that liposome characteristics can be predicted based on process control parameters.

The authors first investigated how liposome size can be controlled during the microfluidic manufacturing process. The convenient computer control on the NanoAssemblr Benchtop allowed them to easily adjust the total flow rate (TFR) between 0.5 mL/min and 2 mL/min and the flow rate ratio (FRR) of the aqueous:solvent phases between 1:1 and 5:1. They found that liposome size decreased as the FRR increased, but was unaffected by increasing the TFR. This showed that liposome size can be controlled by adjustments to the FRR during the manufacturing process. Liposome size, zeta potential, and polydispersity remained essentially unchanged when increasing the TFR. This indicated that microfluidics systems can manufacture liposomes at higher volumetric flow rates and production outputs, which is a key advantage that microfluidics has over other manufacturing methods. The effect that adjusting the FRR and TFR settings had on the lipid composition of liposomes prepared using microfluidics was also investigated. The authors recovered over 87% of the initial lipid amount for all formulations, which suggests that lipid content remains independent of flow rates and flow ratios in the NanoAssemblr Benchtop and confirms the suitability of microfluidics for producing small liposomes with consistent lipid compositions.

The authors also investigated how adjustments to the FRR and TFR settings during the microfluidic production of liposomes affected the in vitro transfection efficiency of plasmid lipoplexes. They made liposomes at different FRR and TFR, which were then complexed with a luciferase plasmid to create a lipoplex and used to transfect cells. They then assessed transfection efficiency by measuring luciferase activity against cells transfected with a Lipofectin™ control. While liposomes prepared at a aqueous:solvent FRR of 3:1 were found to give the highest transfection rate, changes in the TFR did not significantly alter transfection efficiency. The toxicity of these formulations was also tested to verify that transfection efficacy was independent of cell viability and toxicity and neither TFR nor FRR were shown to affect the cell viability. Liposomes produced using microfluidics on the NanoAssemblr Benchtop were found to be non-toxic and easy to manufacture through a robust and reproducible process, which are characteristics required in an ideal transfection vector.

Finally, the authors wanted to understand the mathematical relationship between the TFR and FRR used in the microfluidics process and the liposome size, polydispersity, and transfection efficacy to facilitate the development of an optimized liposome product. Response surface modeling in a design of experiments study confirmed the existence of a relationship between FRR and liposome size, polydispersity, and transfection efficacy while TFR only affected liposome size and production throughput. This indicated that adjusting FRR in the microfluidic mixer can be targeted to obtain desired liposome characteristics, whereas adjusting TFR mainly only changes production throughput. Multivariate data analysis was also used to confirm the significance of FRR in the microfluidics process for the formation of liposomes, indicating that FRR is the crucial parameter to optimize in a formulation to produce liposomes with a desired size, polydispersity, and transfection efficacy.

This paper confirms the advantages of microfluidics as a bottom-up liposome manufacturing method and reveals how statistical analysis can efficiently be used to model and predict liposome size, polydispersity, and transfection efficacy as a function of the process variables during microfluidic mixing. The authors anticipate microfluidics will become the method of choice for liposome manufacturing in the future and that statistical based process control and optimization tools will enhance the reproducibility in this process. The ability of statistical analysis to correlate the relationships between process parameters to predict their effects on liposome characteristics, will lead to a desired and robust product quality that, combined with the ability to rapidly make different formulations on the NanoAssemblr™, can then be used to synthesize and explore a vast range of liposome formulations.

***Note: Current NanoAssemblr® devices operate at higher TFR than the devices used in this study for increased throughput. Using these TFR values with current designs will not produce the same results.

Abstract:

Microfluidics has recently emerged as a new method of manufacturing liposomes, which allows for reproducible mixing in miliseconds on the nanoliter scale. Here we investigate microfluidics-based manufacturing of liposomes. The aim of these studies was to assess the parameters in a microfluidic process by varying the total flow rate (TFR) and the flow rate ratio (FRR) of the solvent and aqueous phases. Design of experiment and multivariate data analysis were used for increased process understanding and development of predictive and correlative models. High FRR lead to the bottom-up synthesis of liposomes, with a strong correlation with vesicle size, demonstrating the ability to in-process control liposomes size; the resulting liposome size correlated with the FRR in the microfluidics process, with liposomes of 50 nm being reproducibly manufactured. Furthermore, we demonstrate the potential of a high throughput manufacturing of liposomes using microfluidics with a four-fold increase in the volumetric flow rate, maintaining liposome characteristics. The efficacy of these liposomes was demonstrated in transfection studies and was modelled using predictive modeling. Mathematical modelling identified FRR as the key variable in the microfluidic process, with the highest impact on liposome size, polydispersity and transfection efficiency. This study demonstrates microfluidics as a robust and high-throughput method for the scalable and highly reproducible manufacture of size-controlled liposomes. Furthermore, the application of statistically based process control increases understanding and allows for the generation of a design-space for controlled particle characteristics.