Development of a methodology to determine drug distribution in polymeric thin film formulations using hyperspectral image analysis

Abstract

This dissertation is focused on the development of analytical methods to determine drug distribution throughout polymeric films using hyperspectral image analysis. The NIR-CI and Raman mapping techniques have been used to analyze the distribution and quantification of drug in a novel pharmaceutical formulation. This pharmaceutical formulation was developed with the goal of maintaining the drug with a specific particle size in a non agglomerated form and to satisfy two commonly encountered pharmaceutical needs: enhanced dissolution rate of poorly soluble drugs and the content uniformity of drugs administered in low doses. In these films, the active ingredient is a poorly soluble drug, which is dispersed in the polymer and additionally; surfactant and lubricant are added. The water poorly soluble drugs have to be uniformly distributed in a film formulation to have an acceptable drug content uniformity. This is the reason why drug distribution is an important factor in these type of pharmaceutical formulations. In order to obtain the desired distribution it is necessary to find those areas of the film where the drug is agglomerated, and use this information to improve the process. Chapter 3 presents the results of the first methodology developed. The first methodology was based on determining the pixels of maximum intensity value at 2080 nm. At this wavelength, pixels with positive values correspond to the drug. A film with large agglomerates was used to develop this methodology. Results show that the drug should be highly agglomerated to find pixels of pure drug. Therefore, the methodology was developed to evaluate the distribution of those pixels that are composed mainly of drug but also containing others components. Agglomerations of these pixels are called drug rich areas in this dissertation. The procedure followed to evaluate these drug rich areas is: to Identify of drug rich areas, observe the distribution of drug rich areas throughout the film surface, and do a visual comparison between different areas or films. This methodology was applied in the images analysis of the chapter 5-7. Chapter 4 compares the previously developed methodology with a new approach using Multivariate Image Analysis (MIA). The previously developed methodology is based on observing the distribution of drug rich areas at 2080 nm. The new methodology is based on the Bharati and MacGregor approach for incorporating the textural information of the image. MIA was used to explore the spectral and spatial relationship between the API and the different excipients. Results obtained with the two methodologies are comparable in terms of drug distribution. Chapters 5 -7 show the results obtained by applying the methodologies developed. These results are based on analysis of high intensity values at 2080 nm and score images. Nanosuspensions produced from wet stirred media milling (WSMM) were used to prepare these polymeric films. In these chapters the effect of stabilizers on controlling growth and agglomeration of the drug, the influence of the drug molecule on the distribution of drug rich areas and the impact of the drying process in the agglomeration of these drug rich areas, were evaluated. Chapter 8 shows the results obtained during the internship. The objective of this internship was to develop skills in solving challenging problems in both fundamental and applied research. This internship was conducted in a Pharmaceutical Chemical Plant during a period of 8 months. During this time, two NIR methods were developed in an effort by the Pharmaceutical Chemical Plant to implement techniques of analysis faster and cost effective. A feasibility study using Raman Spectroscopy for the ID of raw materials also was completed. Chapter 9 summarizes the scientific contribution of this dissertation.