SD consists of three steps of operation: atomization, dehydration, and powder collection. As shown in Fig. 1A, the protein solution feed is sprayed by an atomizer into a drying chamber. Aided by the large specific surface area of the droplets and the hot air, dehydration takes place in a matter of seconds in the laboratory-scale drying chamber. Finally, the dry particles are carried into the cyclone and settle in the product collector. The most important process parameters include drying air inlet temperature (Tinlet), drying air outlet temperature (Toutlet), drying air flow rate (2DA), and atomizing air flow rate (2AR). Of these parameters, Toutlet is the dominating factor in controlling the drying rate and important particle characteristics, such as particle shape and moisture content (see Subheadings 3.2.4. and 3.2.5., respectively). Detailed analysis on stress events associated with SD are described in Subheading 3.2.3. Owing to the importance to process understanding, process and scale-up considerations are delineated in Subheadings 3.2.1. and 3.2.2. Moreover, to explore an emerging technology as a viable alternative to SD, supercritical CO2-assisted bubble drying will be depicted in Subheading 3.2.6.
Many researchers studying SD of proteins have used bench-top spray dryers (e.g., Buchi dryers) for their ability to process a relatively small quantity of the protein—an important attribute if the availability of high-valued proteins is limited. Unfortunately, powder collection efficiency for this kind of dryer is relatively low at 20-50% (8,9). For formulation scientists, it is hard to realize that with appropriate modifications (Fig. 1B), this bench-top dryer can be an efficient and useful manufacturing tool, at least for early phase clinical trials (see Note 1).
Yet, there is another limitation with the bench-top spray dryer—generating particle size of primarily less than 10 | m—the reason why dehydration requires only a few seconds. Certainly, larger dryers offer longer drying times for larger droplets to be dehydrated, but this comes at the expense of the small batch-size capability. To overcome this limitation, a custom-made spray dryer can be designed with a long, narrow drying chamber and a direct in-line filter collection system. The utility of a filter collection system also substantially benefits powder collection efficiency. Such a custom-made spray dryer can produce particles as large as 50 | m in median diameter.
Although SD has evolved into a mature technique for industry-scale production up to a few tons per day (10), within the relatively young biotechnology industry,
the production of high-valued protein powders has been limited to the laboratory bench-top. There is very little information available for considering the scale-up issues of this powder formation process. However, an obvious challenge for the direct application of an industry-scale spray dryer is that such dryers might be prohibitively expensive because of the high-material costs. Again, scale-up strategies are available for the laboratory-scale spray dryer (see Note 2).
A typical concern with SD of proteins/peptides is how thermally labile proteins can resist heat denaturation by hot air. This is a simple physical chemical phenomenon. During the early stage of drying, where the droplet surface remains moisture saturated (i.e., 100% relative humidity), the droplet surface temperature maintains at the wet-bulb temperature that is significantly lower than the hot air temperature. As drying continues, the droplet temperature begins to rise as water diffusion to the droplet surface cannot maintain 100% moisture. At this stage, the protein is primarily in the solid state, and the surrounding air temperature also decreases significantly due to moisture uptake. Thus, thermal denaturation is not typically observed in SD. However, it is a good practice to use a lower inlet air temperature to reduce the potential thermal stress to the protein. Nevertheless, as in FD, protein denaturation often occurs during dehydration; thus, it is necessary to incorporate a stabilizer (e.g., sugars or amino acids) into the protein formulation (11,12). For some proteins, SD can alter the secondary structure (a-helix, P-sheet, and random coil; 13) as observed in lyophilization (14,15), which may irreversibly inactivate the protein. These alterations are attributed to the removal of hydration water molecules that are required to form hydrogen bonds to stabilize the protein's secondary structure (16). Therefore, in developing a biochemically stable spray-dried protein product, it is judicious to dry the protein with a substance (e.g., sucrose or trehalose) that serves as a good water-replacing agent (17,18).
Atomization and the air-water interface are the two other possible sources of stress during SD. It has been demonstrated previously that proteins can sustain shear rates as high as 105 s-1 (19). Mathematical modeling estimates that the shear rate arising from atomization is in the range of 104-105 s-1 at most; therefore, it should not be a significant stress to the protein. However, when shear stress of this magnitude is combined with air-water interface, it may cause significant aggregation for air-water interface sensitive proteins, such as rhGH, bovine serum albumin (BSA), and lactate dehydrogenase (LDH) (20-22).
The structure of most proteins is more or less amphiphilic, i.e., surfactant-like structure. The protein molecules tend to be adsorbed to the air-water droplet interface, where unusual surface energies may cause the protein to unfold, exposing hydropho-bic regions. The unfolded protein may then undergo aggregation by the interaction of the exposed hydrophobic regions with other unfolded molecules until precipitation occurs (23,24). This kind of surface denaturation has a great influence on spray-dried proteins because atomization generates fine droplets with an extremely high-specific surface area (A), that follows the relationship of A = 6/Ddroplet (e.g., 6000 cm2/cm3 for 10-|im droplets). A linear relationship between rhGH aggregation and 1/Ddroplet suggests that aggregation is dominated by the total air-water interfacial area (25).
Three options were found to be effective in minimizing the rhGH aggregation: the addition of a surfactant to prevent the formation of insoluble aggregates, addition of divalent zinc ions to prevent the formation of soluble aggregates, and increasing the rhGH concentration in the liquid feed.
3.2.4. Shape/Morphology of Spray-Dried Particles
Two factors predominate the shape and morphology of the particles upon SD: the rate of droplet evaporation and formulation composition. The speed of solvent evaporation dictates particle quality. Fast drying may cause deformed or defective particles, but slow drying may result in particles that are too wet and sticky to be collected effectively. Meanwhile, spray-dried composition (e.g., materials) prescribes the shape of the spray-dried particles. For example, some materials tend to form solid spherical particles, whereas others form hollow, deformed (shrivelled and cenospherical) or disintegrated particles (26). A hypothetical film, formed at the external surface of the droplets during drying, offers a possible explanation for both types of observed particles. Theoretically, if formed, such a film will encumber the outward diffusion of water and cause the water vapor pressure inside the droplet to increase. At a critical pressure, the film bursts, deforming the particle shape from its original sphericity. Certainly, the extent of such hindrance in diffusion will be dictated by the film properties, such as flexibility, mechanical strength, porosity, and so on.
Without doubt, formulation composition (i.e., the protein and the excipients) determines the film properties which, in turn, governs the observed particle morphology. However, the nature of the protein appears to have a stronger influence on particle morphology because of the protein's sensitive nature to the air-water interface. rhDNase tends to form spherical particles with a smooth-surface morphology (Fig. 2A). Recombinant human anti-IgE antibody typically forms donut-shaped or dimpled particles (Fig. 2B). However, proteins like rhGH and BSA, often form particles of raisinlike morphology (Fig. 2C).
The rate of droplet evaporation may also affect the film properties at the droplet' s surface. A fast-drying rate will promote the formation of a more viscous film at the earlier phase of drying. Because water pressure increases rapidly inside the droplet, the film is prone to burst; therefore, the fast-drying conditions result in a large fraction of the particles containing dimples or holes. For example, based on the notion that a higher outlet temperature prompts a faster drying rate, rhDNase was spray-dried using two outlet drying air temperatures: 46°C and 88°C. Indeed, the particles collected with the outlet temperature adjusted to 88°C had extensive holes, whereas particles collected with an outlet temperature of 46°C were more spherical (Fig. 3).
3.2.5. Factors Affecting the Powder's Residual Moisture Content
The impact of moisture and temperature on solid-state protein stability is well-documented (27,28). Generally, a protein's chemical stability decreases with increasing moisture in the solid because of changes in either the dynamic activity or conformational stability of the protein or from water that serves as a reactant and/or medium for mobilization of reactants (28). In another perspective, residual water can serve as a plasticizer to lower the T of the solid formulation—a physical attribute
unfavorable to protein stability. Furthermore, the level of moisture in the powder may affect particle size and promote excipient crystallization during long-term storage, thereby altering the properties and characteristics of the original powder. Thus, in the developmental stages of a spray-dried powder product, it is essential to understand how the SD conditions affect the powder's residual moisture content. By vapor pressure and mass transfer analysis, the effect of SD conditions on the powder's final moisture content can be elucidated (see Note 3). Overall, to reduce moisture content of the spray-dried powder, escalating the outlet temperature of the drying air is the most effective approach, but it is limited as a result of the possibility of thermally denaturing the protein.
Another important concept about maintaining the dry powder' s moisture content is the humidity control on the powder-processing environment. If uncontrolled, residual moisture stored with the powder following SD may result in protein destabilization.
As an emerging particle formation technology, this innovative method enhances the ability of atomization (or nebulization/aerosolization) and dehydration posed by conventional SD (29,30). Instead of establishing fine droplets by a two-fluid or pressurized atomizing mechanism, the protein/excipients-containing liquid formulation is mixed with supercritical CO2 in a low dead-volume tee, and subsequent expansion out of a capillary flow restrictor into a drying chamber forms a fine emulsion, consisting of microbubbles (supercritical CO2) and microdroplets. Thus, this technology is often referred to supercritical CO2-assisted nebulization by a bubble dryer (CAN-BD).
There are two driving forces attributed to such emulsion formation. These are the sudden physical dispersion of the liquid solution by the rapid expansion of compressed CO2 in the drying chamber and further breakup of microdroplets because of the sudden release of CO2 dissolved in the liquid solution during their intimate contact in the mixing tee. Indeed, these physical mechanisms render added power to aerosolization, thereby readily producing a fine powder of less than 5 |im in diameter.
Owing to the large-specific surface area, the microdroplets can undergo rapid dehydration even using a low-temperature drying gas, e.g., dry nitrogen of 25-65°C. Certainly, such a drying temperature is much lower than that encountered in conventional SD, normally at 80-150°C. Therefore, the thermally induced stress by SD can be mitigated, which is one prominent advantage offered by CAN-BD.
There is a stress event specifically associated with CAN-BD, i.e., the interactions between protein and CO2, occurring at the interface of the microdroplet and supercritical CO2 and/or within the microdroplet where the protein encounters the dissolved CO2. Although the number of proteins tested by CAN-BD or supercritical CO2 methodologies in general remains low, several case studies have been reported that demonstrated the overall effect on protein stability is mild if the proteins are formulated with a proper buffer and pH, an amorphous disaccharide (e.g., trehalose or sucrose), and surfactant (e.g., Tween-20 or Tween-80). These proteins include LDH (29), trypsinogen (30,31), anti-CD4 (31), and a-1-antitrypsin (31).
Overall, CAN-BD is a viable technology capable of generating fine powders that greatly benefit formulation scientists in processing proteins of limited availability in early development work. Equally important is the demonstration of CAN-BD for powder manufacturing at a commercial scale (32). Yet, a broader use of this method for producing protein powders for various biopharmaceutical and drug delivery applications requires a more extensive compilation of protein case studies and the ability of particle size manipulation to go beyond the 5-pm barrier.
Was this article helpful?