Abstract:
The aim of the current work is the enhancement of the solubility of a poorly soluble drug using albumin nanoparticles by applying quality by design approach. using several tools as artificial neural network and response surface methodology for the optimization of albumin nanoparticles. Silymarin is an extract of dried milk thistle seed with many uses such as anticancer drug, anti-inflammatory, antioxidant and hepatoprotective agent. One of the main problems in the Silymarin that it has poor water solubility and poor bioavailability. This problem could be solved by its formulation in albumin nanoparticles.
A complete quality target product profile has been constructed, and Ishikawa diagrams were very beneficial in the risk assessment study. Fractional factorial design was used in the screening, where time of stirring, albumin concentration, pH, drug amount, amount of ethanol and the type of solvent were the critical process parameters/ material attributes (CPP/MA), and were tested on the particle size, polydispersity index and the encapsulation efficiency, which were considered as the critical quality attributes (CQA). Whereas, D-optimal design the response surface design (RSD) and was used for the optimization step, where the drug amount and the albumin concentration were only tested on the same previously measured CQA.
Artificial neural network (ANN) was applied, by taking the CPP/MA of the optimization design as the inputs and the measured CQA as the outputs, where the obtained correlation coefficients were compared with that obtained from the RSD,, and was found to be higher than the RSD. Design space and control strategies were generated, where an optimized formula as suggested from software was prepared and its results were compared with the expected ones to calculate the % bias, where the results reasonable agreement indicating the validity of the design.
Thus quality by design was found to be a successful approach in the formulation and optimization of albumin nanoparticles loaded with silymarin. ANN showed better results than RSD, which could be used in many applications in industrial formulations.
