Poorly soluble small molecules drug candidates continue to constitute a significant portion of the development portfolio [1,2] and over the last decade amorphous solid dispersion technology has taken on an increasingly important role in enabling poorly soluble molecules transition from pre-clinical evaluation to clinical development and beyond to drug approval [3,4].
Spray Dried Dispersions (SDDs) require special attention during development and scale up because of the potential interrelationship of factors that affect crystallization tendency, dissolution, formulation and process conditions. Scheme 1 provides an overview of the important conceptual process development segments required for development of an SDD that is eventually provided as an oral tablet for the final dosage form.
Integrated understanding of ASD characteristics is needed for efficient development and a range of characterization techniques are applied across development stages [6,7]. Particulate based characterization techniques such as PLM, SEM, PSD are helpful for assessing properties of ensembles of particles, but are limited in the level of detail they can provide in a time and cost effective manner. Furthermore, they don’t yield a quantitative picture of the interior and exterior microstructural characteristics of the ASD particles which can potentially impact performance behavior and may need to be controlled at a later stage to optimize downstream processing characteristics such as compaction.
Within the last several years, application and refinement of techniques using correlative, three-dimensional (3D) X-Ray Microscopy (XRM) and Focused Ion Beam – Scanning Electron Microscopy (FIB-SEM), have enabled direct imaging of the internal microstructure of drug products at a representative volume at high resolution. XRM in particular is a non-destructive technique that can characterize a variety of pharmaceutical product systems at different stages of manufacture, under different stability conditions or at different release times. The combination of these imaging technologies allow non-invasive and facile visualization of the interior characteristics of pharmaceutical composite materials, upon which further quantitative analysis can be performed. Because SDDs may need to be transferred to/from different production sites/companies/scale and characteristics may be required to be duplicated, understanding of the critical aspects that affect the SDD’s behaviors are very important and microstructural investigation can yield actionable information in this regard.
Assessment of 3D particle and domain distribution is a useful means for rigorously comparing the attributes of particles made by different processing methodologies. Figure 2 compares slices of 3D images of ASDs of equivalent composition made from either a spray drying process or ultrasonic mechanical dispersion . The images clearly show differences in particle size, but also reveal that the mechanically formed dispersion (Fig 2b) contains a proportion of large particles that are porous and a portion of smaller particles similar in density to those made in the spray drying process. The size distributions for the two samples also show the difference in overall size but because XRM can selectively quantify a differentiable component, the ASD-like particles within the mechanical dispersion samples can be compared to that of the spray dried powder. Once XRM is conducted a range of analysis can be conducted which can yield quantification of pore size, surface area, component domain size, as well as other particle morphological attributes.
Figure 3 highlights the key analytical results obtained from two solid spray dried particles  from identical formulations spray dried under different conditions (drying gas:feed flow rate ratio; outlet temp). The two SDDs showed significantly different dissolution profiles despite appearing visually the same in SEMs and having comparable other characterization data (Malvern PSD, specific surface area, moisture sorption, thermal analysis, contact angle). Because of the resolution and 3D versatility of the XRM digital data set, the wall thickness distribution profile and other internal microstructural characteristics could be determined and compared (Fig 3). The P2 batch containing spray dried particles with thicker particle walls and large bee-hive morphologies resulted in lower extent and rate of dissolution. The image based calculated exterior surface area for batch P2 was lower than that of batch P1. Based on the combination of results, the cause of poorer tablet dissolution of the P2 batch was inferred to be due to the thicker particle walls and lower dissolution available surface area. When formed into tablets, those features resulted in less pore connectivity and smaller tablet pore size compared to tablets made from P1, which in turn negatively impacted the dissolution. This second example demonstrates the value of XRM analytics in establishing root cause understanding where the typical characterization tools are not able to provide a direct correlation with performance behavior for process optimization.
Powder and particle properties of SDDs are of particular importance in later stage development because these properties can significantly affect downstream manufacturing [5, 10] where powder flow and/or compression characteristics can influence dosage form manufacturability and performance. SDDs often require high loading in tablet formulations, consequently ensuring good and consistent compaction properties are important for making robust tablets. Physicochemical characteristics of the dried SDD particle will be influenced by a range of process operating conditions in a dynamic fashion [11,12]. Particle size distribution by laser light scattering is routinely applied but does not capture all potential functional particulate characteristics. Qualitative morphology assessment by visual inspection of SEM or some level of 2D image analysis can detect raisin vs hollow sphere morphologies that result from variation in spray drying conditions when making SDDs. However, these methods are unable to easily reveal aspects of the interior of the particle, such as wall thickness vs. size relationships and be performed on large numbers of particles. Deeper 3-D analysis using uXRT can obtain a richer set of data that can be utilized to more clearly identify the relationship between process parameters and physical properties.
Figure 4 shows how a digital imaging workflow was incorporated for studying the process parameters (Tout; spray drying outlet temperature) for producing the SDD . XRM characterized both the particle size and interior wall thickness as well as exterior shape quality of the particles (bubbles vs raisins). This quantitative information could then be correlated to the tableting performance data to optimize compaction related tablet tensile strength. Compared with conventional techniques, 3D image-based analysis results were more sensitive to the spray drying process parameter changes. They clearly demonstrated a positive correlation between PSD and Tout and a negative correlation between Tout and the envelope density or wall thickness. Interrogating both particle morphology and internal microstructure of ASDs thus provided insight into how the particle properties influence downstream processability. It has also been shown recently that XRM based microstructural determination of the solid volume fraction of the spray dried particle is a key parameter that could be utilized in modeling spray drying process formation kinetics [14,15].
The application of uXRT and advanced image analytics processing capabilities to interrogate both the ASD as drug product intermediate and in the finished dosage form allows greater insight into factors related to QbD and establishment of quantitative control strategies for SDDs. The examples demonstrate that microstructure analysis provides useful and detailed information on material attributes that conventional characterization typically cannot access. uXRT can provide value-added linkages between process parameters and particulate quality attributes and contribute to root cause assessment. Because uXRT analysis can be efficiently applied to large numbers of particles contained in a unit dose, it can be applied across the spectrum of development for SDDs, from initial formulation and process comparison, to providing feedback on design space and scale up where it can reveal critical formation-based factors that may impact manufacturability and in vitro performance.