Rapid fire talk
Purpose: Microstructures are at the foundation of optimizing drug formulations and release performance. From the particle size of active ingredients to microporosity, these factors are critical towards understanding release mechanisms and engineering an effective drug product. To characterize these interconnected microstructures, an analytical technique must have high resolution, sensitivity, representative sampling, and 3D visualization. Conventional techniques are lacking in one or many of these areas. Digital drug formulation (DDF) is a novel platform that implements high resolution 3D imaging coupled with artificial intelligence (AI) analytics to quantify and understand the impact of microstructures on drug release.
Methods: The DDF platform utilizes microstructure data obtained by 3D imaging modalities, including x-ray microscopy (XRM) and focused ion beam scanning electron microscopy (FIB-SEM). Through a cloud computing system DigiM I2S, image data is constructed into a digital sample of the drug product or intermediate. This digital twin of the real sample can be continually reused and reanalyzed. The microstructures are identified and labeled through AI-based analysis. A matrix of quantitative descriptors including volume fractions, particle size, particle dispersion, porosity, and transport properties are computed. These descriptors can be numerically adjusted to generate new drug microstructures. The behavior of real and generated microstructures on drug release is then studied with image-based release prediction.
Results: The DDF workflow has been applied to study the sensitivity of drug release from a PLGA microsphere by changing the microstructure and evaluating image-based release profiles. Comparing the real microstructure with numerically generated microstructures, a significant impact on drug release was observed. By removing porosity, a release period of 3 weeks was extended to beyond 30 days. This evaluation provided guidance to formulation scientists on the longest drug release possible with this microsphere platform.
Conclusions: The DDF model provides a framework to study the sensitivity of microstructures to formulation parameters (e.g. polymer choice, particle morphology) and process conditions (e.g. temperature, compaction force). The DDF platform can generate numerical drug formulations, evaluating the impacts of drug loading and micro-porosity. The numerical model allows a formulation scientist to rapidly traverse multi-variable parameter space, narrowing down an optimal formulation and best processing conditions.