(W0930-02-13) Continuous Manufacturing of Injectable Nanoemulsions: Methods and Characterization

Purpose: The goal of this work is to develop continuous and robust methods for manufacturing of injectable nanoemulsions. Nanoemulsions have been used as drug delivery systems across indications, including cancer, eye disease, vaccines and anesthesia for several decades[1]. Conventional manufacturing of nanoemulsions is typically a batch process, often resulting in failing batches, difficulty to scale up[2] and long manufacturing times. Continuous manufacturing promises to improve the processes and the resulting products by precisely controlling the mechanisms that lead to droplet formation and eliminating manufacturing steps that cause variability. Selection of appropriate characterization methods is key for process optimization. In this work, particle sizing, stability, ability to filter sterilize the nanoemulsions and process efficiency were the metrics used to compare the Continuous and conventional manufacturing methods.

Methods: >A small molecule API was encapsulated in a O/W nanoemulsion at concentrations 5-7 mg/ml. This particular API is practically insoluble in water, with solubility < 0.01 mg/ml. Soybean oil, egg phospholipids and API were the key ingredients in the nanoemulsion, with the oil phase being about 28 wt % of total. Nanoemulsions were manufactured using a novel Continuous emulsification method and the conventional, batch method. Both methods rely on high pressure homogenization (HPH), which is a processing method with high energy rates, of about 10⁸ -10⁹ W/kg[3]. However, the conventional method includes a discrete premixing step prior to HPH, in which the oil and water phases are first mixed to form a quasi-stable coarse emulsion. The coarse emulsion is often characterized by variability in terms of particle size and stability. Such a variability may negatively affect the properties of the final nanoemulsions. A rotor stator mixer with typical energy rates of 10⁵-10⁶ W/kg[3], was used for the formation of the premix. In contrast, Continuous emulsification does not include a discrete premixing step. The method  includes an in-line static mixer which is positioned at the inlet of the HPH. The water and oil phases form continuous streams and  are mixed as they flow to the inlet of  the high pressure homogenizer. Typical energy rates of the in-line mixer are substantially lower than those of the rotor-stator mixer, and are in the range of 10² -10⁴ W/kg. In both methods, the nanoemulsions were passed through the HPH several times until the appropriate particle size was achieved. Nylon 0.22 micron filters were used to filter sterilize the nanoemulsions. Dynamic Light Scattering was used to measure the average particle size of the nanoemulsions, while analytical centrifugation was used to determine possible instability mechanisms and measure stability parameters.  Analytical centrifugation was performed using a LUM gmb instrument.  Samples of nanoemulsions are exposed  to centrifugal forces, in this case 2,300 g for 150 minutes, at 21 ^oC. Temporal and spatial light transmission signals from the samples correlate with migration of particles and/or with particle growth, and therefore assess the emulsion stability.

Results: Figure 1 shows the average particle size of the nanoemulsions as a function of the number of passes through the HPH  for the Continuous and conventional methods. It can be seen that Continuous emulsification consistently resulted in smaller particle size than the conventional method. With Continuous emulsification, the target of 85 nm is achieved after two passes. However, with the conventional method, this target was barely achieved after eight passes. Both methods resulted in particle size distributions with similar polydisperivity indexes (PI) in the range of 17-22. Continuous emulsification resulted in nanoemulsions that were filter sterilizable at all steps of processing. However, with the conventional method this was possible from the second pass on. Figures 2a and 2b show the temporal and spatial transmission profiles of nanoemulsions from the analytical centrifugation. Higher overall transmissions and smaller differences between the initial and final profiles indicated that Continuous emulsification resulted in nanoemulsions with smaller particle size and higher stability when compared  to nanoemulsions from the conventional method. Figure 3a shows the separation rates of the nanoemulsions in %/min during 10-40 minutes of analysis. Figure 3b shows the percentage of nanoemulsions that separated during the same time.  Continuous emulsification resulted in separation rates 3.5-10 times lower and percentage of separated nanoemulsion 1.6-4 times lower than those from the conventional method.

Conclusion: Continuous emulsification was found to be over 400% more efficient than the conventional batch method, and resulted in nanoemulsions with substantially smaller particle size and better stability characteristics. The target of 85 nm was achieved after two high pressure homogenization passes with Continuous emulsification, while it was barely reached after 8 passes with the conventional method.  Stability parameters, such as separation rates and percentage of material separation were substantially lower in nanoemulsions produced with Continuous emulsification, indicating superior stability of those nanoemulsions. Furthermore, the elimination of the premix step, increases the overall efficiency of Continuous emulsification even further in terms of processing times, energy demands and capital equipment costs.

References: 1. R. J. Wilson, Y. Li, G. Yang and C. Zhao (2022), “Nanoemulsions for drug delivery”, Particuology Volume 64, Pages 85-97, https://doi.org/10.1016/j.partic.2021.05.009.
2. Hakansson, A. (2017) “Scale-down failed, dissimilarities between high-pressure homogenizers of different scales due to failed mechanistic matching”, J. Food Engg. 195, 31-39.
3. Calabrese, R.V., S. M. Kresta, M. Lui, (2014) “Recognizing the 21 Most Influential Mixing Research”, CEP, January 20-29.

Acknowledgements: The authors would like to thank Mr. Jonathan Doughty for helping us with the production of the emulsions. The authors declare that they have no conflicts of interest.

Figure 1. Average particle size as a function of number of homogenization passes resulting from conventional and Continuous methods. Superimposed is the specification for the average particle size.


Figure 2. Temporal and spatial transmission profiles of nanoemulsions produced with (a) conventional and (b) Continuous emulsification methods after three homogenization passes. Initial profiles are shown in purple, while later profiles are shown in shades of yellow.


Figure 3. (a) Separation rate, and (b) % of material separated during analysis times of 10-20 minutes.