Antisolvent vapor precipitation: the future of pulmonary drug delivery particle production?

Abstract

Pulmonary drug delivery requires particles with a fine and uniform size distribution, 1 – 5 µ hydrodynamic diameter, for targeted delivery and higher uptake within the pulmonary tract. The antisolvent vapor precipitation method is a new alternative method to generate particles with such characteristics. Development so far has shown that the technique is applicable to disaccharides (lactose), magnesium sulfate, hydrophobic and protein particles. The method may also be extended to generate fine encapsulation particles and to produce disaccharides (lactose) and protein particles blend in a single step.

Keywords::

• antisolvent
• precipitation
• spherical particles
• spray drying
• submicron
• uniform

1. Introduction

This editorial describes developments of a new method to make ultrafine and uniform particles for pulmonary drug delivery. Particle size and uniformity are very important characteristics in the delivery of drugs via the pulmonary route [1-3]. Uniform particle size ensures that the aerodynamically transported drug particles are more targeted. This is because the behavior of the particles in following or traversing the inhalation airflow (resulting in deposition) is highly affected by their size and density, which is described by the aerodynamic diameter. Therefore, particle size uniformity will ensure that particles in a blend experiences similar air flow interaction resulting in similar trajectory within the airways. In addition to the need for uniform particle sizes, smaller particles will also enhance the dissolution or absorption rate of the pulmonary delivered drugs [4]. This serves to overcome the mucous layer in the lungs so that the drug materials are absorbed into the bloodstream before being expelled. This aspect has been the subject of many studies, particularly in the delivery of protein or poorly water-soluble drugs via the pulmonary tract. In many cases, reducing the size of the drug particle might be an important option in increasing the uptake of these drug particles.

One common method to generate drug particles for pulmonary delivery, or as a general particle production process in the pharmaceutical industry, is the spray drying technology [5]. Although this process has been the work horse in particle production for the industry, one challenge is in producing uniform particles in economical quantity. This is because conventional nozzle or rotating atomizers typically generate a wide distribution of droplets. In conventional spray dryers, each droplet will dehydrate into a single particle. Therefore, the particle size produced will follow that of the initially atomized droplet. The relatively large initial droplet size is also closely linked with the size and throughput of the atomizer. Although there may be variations in atomizer design and operating conditions, in general, the size distribution may be reduced by using smaller atomizers, which also reduce the overall particle size. This strategy has been utilized in commercial laboratory scale equipments [6]. However, the throughput is relatively low and it might not be economical for routine production purpose. There are other strategies used such as including surfactants and additives which are highly material-specific. This editorial focuses on a strategy to alleviate these aforementioned factors in the spray drying process. The key is in replacing the hot convective drying air with ethanol vapor (antisolvent vapor) as the drying medium.

2. Ethanol vapor as the convective drying medium

The premise of this concept lies in the absorption of ethanol vapor into the droplet during the convective drying process [7]. As a result, particles are then formed by the precipitation process from within the droplet, due to the presence of ethanol, rather than by progressive solidification due to moisture removal by hot air. Figure 1 illustrates the difference between this new concept and the conventional spray drying. Typical spray drying produces a single particle from each droplet, whereas with the antisolvent vapor precipitation, hundreds of smaller particles can potentially be produced from a single droplet. This concept, so far, has been developed using the modified single-droplet drying technique in which a single droplet of 1 – 2 mm in diameter is statically suspended exposed to a convective stream of ethanol-nitrogen medium, mimicking a droplet within a spray dryer [7]. Initial developmental work was undertaken with lactose as a model sugar pharmaceutical. Uniform spherical micron-sized amorphous particles or pollen-like crystalline particles were produced (Figure 2A and B). The difference in morphology is due to different ethanol vapor concentrations. Although the exact detailed mechanism is still far from certain at the moment, there was an apparent critical ethanol vapor concentration from which a higher concentration resulted in the precipitation of the spherical particles [8]. Investigation on the detailed mechanism of the process is still in progress. The hundreds of lactose microspheres in Figure 2A displayed high uniformity with a mean diameter of 2 µ with a span of ±1 µ. This initial observation illustrated the potential of this concept to produce uniform fine particles. Surprisingly, the concept also produced amorphous lactose particles which were not normally obtained by precipitating lactose from ethanol-water mixtures [9]. We believe this is due to the unique mass transfer behavior of ethanol into the droplet, which sets it apart from the usual liquid-based antisolvent technique from which the liquid antisolvents are added to the formulation resulting in particle precipitation.

Figure 1. Comparison between conventional spray drying and the proposed antisolvent vapor precipitation approach is given.

Figure 2. Antisolvent vapor precipitated particles: (A) lactose microspheres; (B) lactose micro crystals; (C) lactose + fish oil; (D) magnesium sulfate; (E) lactose + magnesium sulfate; (F) whey protein isolate microspheres and (G) whey protein isolate + lactose segregated particles are shown.

In view of the capability to produce uniform microsphere, the possibility of extending this concept to make micro-encapsulates for hydrophobic materials was further explored. Most encapsulates dealing with dispersed hydrophobic materials tend to form a continuous solid matrix as a large particle. Exploratory experiments undertaken showed that mini-encapsulates can be formed within a single droplet with the encapsulating materials [10], in this case lactose, forming fibrous spheres directly onto the emulsified oil material (Figure 2C). This illustrates the potential of the technique to be an alternative to form micron-sized encapsulate particles. At the moment, however, the size of the micro-encapsulate is still relatively large and it was difficult to control the size of the distribution as it was highly linked with the size of the emulsified hydrophobic materials. In treating acute asthma, magnesium sulfate can be used as drugs [11,12]. With the antisolvent vapor precipitation technique, micron-sized uniform magnesium sulfate particles can be produced albeit with slightly wrinkled surface morphology (Figure 2D). Combining the magnesium sulfate with dissolved lactose produced similar uniform particles but exhibiting a smoother surface morphology (Figure 2E). It will be interesting to further extend the technique to produce other pulmonary drug particles.

Further developmental work was undertaken for protein-based particles using whey proteins [13]. Uniform particles of even smaller sizes were produced, down to the scale of hundreds of fine particles with a mean size of 200 nm (Figure 2F). A different strategy was required for producing such protein particles which involved the adjustment of the initial pH of the droplet. Although the complete dissolution of these particles is not fully assessed in the preliminary exploratory work, dispersion of the protein particles was observed from the single droplet dissolution study. From the experimental evidence, we believe that the technique can be used in producing protein-based particles for pulmonary delivery. The capability to generate submicron (nanoscale particles) will offer a useful avenue for improving the dissolution and uptake of potential protein-based pharmaceutics [14,15]. It is noteworthy that the uniformity, however, is relative to the length scale of the particles produced. In the micrometer length scale, as observed for the lactose microspheres, the variation of ±1 µ in size can be considered relatively ‘uniform’, when compared to typical spray-dried particles. On the other hand, in the submicron length scale, particle sizes produced can vary by hundreds of nanometers as illustrated in Figure 2F. It is unclear, at the moment, if optimization of the process can narrow the size distribution of the particles in the submicron scale.

Another question then arises: What happens if we mix protein- and sugar-based excipient material into the initial droplet? Significant segregation can then occur as illustrated in Figure 2G. Such particles resemble a combination of protein particles embedded in a matrix of crystalline lactose. If further broken down, one possible method is via the ultrasonic technique; such morphology will resemble blended excipients and protein particles. Although not shown here, both the segregated materials can also be produced into spherical microspheres by adjustment of the antisolvent vapor concentration. Therefore, although the direct control in the formation of such morphologies is still far from certain, there is potential to generate such blends in a one-step approach during the particle formation process.

3. Conclusion

This editorial discussed the development of the antisolvent vapor precipitation technique for potentially producing particles for pulmonary drug delivery. The idea is to replace the hot air in a spray dryer, a common process used to produce drug particles, with ethanol (antisolvent) vapor. High uniformity and the fine size of the particles produced will certainly be attractive for pulmonary drug delivery application. From the lactose experiments, amorphous or crystalline morphology can be produced with the former being capable of enhancing dissolution of the delivered particles. The concept can also be extended to make micro-encapsulates for hydrophobic materials and was effective in producing magnesium sulfate microspheres – the latter is used in the treatment of acute asthma. Even finer submicron particles can be produced for protein-based materials which can potentially enhance the uptake of these materials in the pulmonary tract. There is possibility to use this approach as a one-step process to produce sugar-based protein particle blends.

4. Expert opinion

The developments in the preceding section were undertaken with the single droplet experimental technique which uses relatively large droplets of 1 – 2 mm in diameter. A typical droplet size in an actual spray dryer ranges in the order of 10¹ – 10² µ. In view of this, further experiments with actual sprayed droplets in the range of 15 – 60 µ statically exposed to convective antisolvent vapor were already undertaken in a series of uniquely controlled experiments. Reports for these experiments are now currently underway, awaiting provisional patenting of the process. In our opinion, the successful production of the fine uniform particles from this latest phase of experiments is clearly a proof of concept of this new production method. The team in Monash University is currently in the midst of implementing this in an actual dryer.

Once this technological hurdle is overcome, larger quantities of powders can then be produced for aerosolization analysis via the Anderson Cascade techniques. It is envisaged that subsequent developments will then shift toward the formulation aspect for this technology. One key aspect to consider would be the required contrasting solubility of the drug material in water and ethanol. This aspect will be even more complex if multiple solutes are present in the droplet. Such a situation might arise if additives are required to stabilize the particles or to generate a blend of particles using the potential one-step approach described earlier. On top of the formulation aspect, the ethanol vapor absorption process will also determine the final morphology of the particles. In view that the technique may produce amorphous particles such as those observed for the lactose particles, stability profile of the particles will be another important consideration in using the technology.

One can question on the economics of this method. Would the usage of ethanol vapor present a less economical option, and also a more hazardous one, when compared to that of hot air? The advantage of this process is the potential to generate fine uniform particles to enhance drug particle delivery where these two parameters (size and uniformity) are important. Therefore, the economic benefit in enhancement of the delivered drug should outweigh the additional costs if it were to be economically viable. Another key aspect is that this process has the potential to make ultrafine particles without the need for very fine droplet atomization which tend to limit the throughput of the process. In addition, the technique utilizes low (close to ambient) pressure in contrast to other high-pressure techniques such as the supercritical fluid precipitation technique [16]. Although lower drying temperatures will have to be used in the presence of ethanol vapor, due to the potential hazards, such a condition might even be more favorable for protein particle production in light of possible thermal denaturation and aggregation associated with typical spray drying processes. These will certainly be important considerations, particularly from the manufacturer's perspective.

Declaration of interest

Financial support was gained from the Australia Research Council under the Discovery Project (DP130104836) to develop the antisolvent vapor precipitation technique.